SDRF-07-P-0019-V1.0.0
Use Cases for Cognitive Applications in Public Safety
Communications Systems
Volume 1: Review of the 7 July Bombing of the London
Underground
Approved 8 November 2007
SDRF-07-P-0019-V1.0.0
SDRF
SDRF-07-P-0019-V1.0.0
ACKNOWLEDGEMENTS
This document was drafted by the Public Safety Special Interest Group of the SDR Forum. In
addition, the document was reviewed by members of the National Institute of Justice
Communications Technology Working Group, and by representatives of the Intelligent
Transportation System United Kingdom who also have experience in law enforcement and incident
command. Their contributions were significant and gratefully appreciated.
-ii-
SDRF
SDRF-07-P-0019-V1.0.0
TABLE OF CONTENTS
1.
Introduction............................................................................................................................. 1
1.1.
The SDR Forum Public Safety Special Interest Group .................................................. 1
1.2.
Document Overview ....................................................................................................... 2
2. Background and Assumptions ................................................................................................ 3
2.1.
Cognitive Radio Technology .......................................................................................... 3
2.2.
Assumptions.................................................................................................................... 5
3. Methodology ........................................................................................................................... 7
3.1.
Select Scenarios .............................................................................................................. 7
3.2.
Develop Timeline............................................................................................................ 7
3.3.
Identify Use Case............................................................................................................ 8
3.4.
Analyze Use Case ........................................................................................................... 8
3.5.
Review Results with Public Safety ............................................................................... 10
4. London Bombings of 7 July 2005......................................................................................... 11
4.1.
Scenario Assumptions................................................................................................... 11
4.2.
Scenario Timeline ......................................................................................................... 12
4.3.
Cognitive Use Case Discussion .................................................................................... 12
4.3.1.
Use Case 1: Network Extension for Coverage and Reachback ............................ 12
4.3.2.
Use Case 2: Dynamically Access Additional Spectrum ....................................... 17
4.3.3.
Use Case 3: Temporarily Reconfigure First Responder Communication Device
Priorities ............................................................................................................................... 21
4.3.4.
Use Case 4: Interface to Non-First Responders.................................................... 25
5. Multi-Use Case Functional Capabilities ............................................................................... 30
5.1.
Roles ............................................................................................................................. 30
5.2.
Command and Control.................................................................................................. 31
5.3.
Interoperability.............................................................................................................. 34
5.4.
Restoring Default Configurations ................................................................................. 34
5.5.
Security ......................................................................................................................... 35
5.5.1.
User and Role Authentication............................................................................... 35
5.5.2.
Device and Network Authentication..................................................................... 37
5.5.3.
Reconfiguration Controls...................................................................................... 37
5.6.
Transition and Interface to Legacy System .................................................................. 37
5.7.
Interaction of Use Cases ............................................................................................... 38
5.8.
Standards....................................................................................................................... 38
6. Summary ............................................................................................................................... 40
Appendix A: London Bombing Scenario ................................................................................... A-1
-iii-
SDRF
SDRF-07-P-0019-V1.0.0
LIST OF FIGURES
Figure 4-1. Map of the London Bombing Incident Locations ..................................................... 11
Figure 4-2. Network Coverage Extension Use Case Example ..................................................... 14
LIST OF TABLES
Table 1. Possible Dynamic Prioritization Approaches ................................................................ 24
Table 2. London Bombing July 7, 2005 Timeline of Events..................................................... A-1
-iv-
SDRF
SDRF-07-P-0019-V1.0.0
EXECUTIVE SUMMARY
This report is the first in a planned series of reports to be written by the SDR Forum to develop
concepts for the application of cognitive radio technology to enhance the communications
capabilities of public safety first responders. The objective of this series is to provide:
•
Researchers and system developers with an understanding of the desired cognitive
capabilities, from which technical requirements and specifications can be derived;
•
Regulatory agencies with an understanding of the regulatory issues and
identification of potential changes that may be required to fully utilize evolving
cognitive radio technology to benefit public safety; and
•
Public safety agencies with an understanding of the potential value of cognitive radio
technology and an understanding of policy and procedural changes that may be
required to fully utilize evolving cognitive radio technology and regulatory changes.
The methodology for developing cognitive use cases is based on an analysis of response to
actual or hypothesized events. This report is an analysis of the events of 7 July 2005, a terrorist
attack involving coordinated explosions of bombs in and around the London Underground. This
analysis is not an evaluation of that response, but instead uses the lessons learned from real events
and the observed response to envision how evolving cognitive technology could enhance the ability
of responders in the future to communicate more effectively and efficiently than available
technology allowed.
Based on the analysis, we conclude that development of cognitive capabilities has potential to
dramatically increase the ability of incident commanders. Four examples of how cognitive radio
technology could be utilized in such situations (defined as cognitive use cases) are identified and
described in this report:
1. Network extension for coverage and reachback.
2. Dynamic access of spectrum.
3. Dynamic prioritization.
4. Dynamic network configuration to include non-first responders.
The cognitive radio functional capabilities, regulatory implications, and policy and procedure
implications for each use case are analyzed as well. Beyond capabilities associated with specific use
cases, additional capabilities that apply to all of these use cases are also identified. This report was
originally drafted by the Public Safety Special Interest Group within the SDR Forum, which
includes representatives of public safety agencies, manufacturers, research & development
organizations, and regulators. The report then underwent additional review and comment by public
safety practitioners in both the United States and the United Kingdom.
Although there are significant technical, regulatory, and procedural developments required to
realize the benefits, development of cognitive capabilities can dramatically increase the ability of
incident commanders. First responders will be assured that critical information will flow as needed
despite changes in coverage, connectivity, and loading on communications systems.
-v-
SDRF
SDRF-07-P-0019-V1.0.0
This page intentionally left blank.
-vi-
SDRF
SDRF-07-P-0019-V1.0.0
1. INTRODUCTION
The maturing of software defined radio technology and evolving concepts of cognitive radios
hold great promise for public safety communications. The Public Safety Special Interest Group
(SIG) of the Software Defined Radio (SDR) Forum has already released a report summarizing
potential applications of SDR and cognitive technology to public safety.1 One of the key areas of
interest defined in that report is cognitive applications.
This report is the first in a planned series of reports to be written by the SDR Forum to develop
concepts for the application of cognitive radio technology to enhance the communications
capabilities of public safety first responders. The purpose of this series of documents is to explore
in greater detail specific examples of how cognitive applications can be used in public safety
communications networks to enhance communications capabilities.
The objectives of the Public Safety SIG in generating this series of documents are as follows:
1. Provide researchers and system developers with an understanding of the desired cognitive
and related functional capabilities, from which technical requirements and specifications can
be derived;
2. Provide regulatory agencies with an understanding of the regulatory issues and
identification of potential changes that may be required to fully utilize evolving cognitive
radio technology to benefit public safety; and
3. Provide public safety agencies with an understanding of the potential value of cognitive
radio technology and an understanding of policy and procedural changes that may be
required to fully utilize evolving cognitive radio technology and regulatory changes.
1.1. The SDR Forum Public Safety Special Interest Group
The SDR Forum is an open, non-profit corporation dedicated to supporting the development,
deployment, and use of open architectures for advanced wireless systems, with a mission to
accelerate the proliferation of SDR technologies in wireless networks to support the needs of civil,
commercial, and military market sectors. Activities focus on:
•
Developing requirements and/or standards for SDR technologies, including working
in liaison with other organizations to ensure that Forum recommendations are easily
adapted to existing and evolving wireless systems;
•
Cooperatively addressing the global regulatory environment;
•
Providing a common ground to codify global developments;
•
Serving as an industry meeting place.
The Public Safety Special Interest Group is one of several special interest groups within the
Forum that bring together developers, users, regulators, and educators to address issues specific to
the application of SDR technology to a particular domain or market area. Goals of the Public Safety
SIG are to interface with the public safety community (including both users and vendors), to raise
1
SDR Forum, Software Defined Radio Technology for Public Safety, Software Defined Radio Forum Report SDRF06-A-0001-0.0, 14 April 2006, available at www.sdrforum.org.
-1-
SDRF
SDRF-07-P-0019-V1.0.0
awareness of SDR, to publicize the activities of the Forum in addressing those issues, and to
increase participation of the public safety community in the SDR Forum. The Public Safety SIG
also interacts with other committees and working groups within the Forum to provide the public
safety community’s inputs into the publications and initiatives undertaken by the Forum. In the case
of this report, members of the SDR Forum Security and Cognitive Radio Working Groups have
participated in the preparation of this report. The Public Safety SIG is a unique venue, because
participation in the SIG has historically included public safety organizations, land mobile radio
vendors, manufacturers of SDR for military applications, software developers, researchers, and
regulators.
1.2. Document Overview
The methodology for developing cognitive use cases is based on analysis of response to actual
or hypothesized events. This report is an analysis of the events of 7 July 2005, a terrorist attack
involving coordinated explosions of bombs in and around the London Underground. This analysis
is not an evaluation of that response. Rather, it is an attempt to use the lessons learned from real
events and a real response to envision how evolving cognitive technology could enhance the ability
of responders in the future to communicate more effectively and efficiently than current technology
allows.
The background and assumptions of our analysis is provided in Section 2. In Section 3, we
provide an overview of the methodology to be used throughout this series of documents.
Identification and analysis of specific use cases is included in Section 4. Section 5 includes a
discussion of issues that apply to multiple use cases. A summary is provided in Section 6, and a
timeline of the events of the London Bombing Scenario is presented in Appendix A.
-2-
SDRF
SDRF-07-P-0019-V1.0.0
2. BACKGROUND AND ASSUMPTIONS
As noted in the Introductory Section, the potential value of cognitive capabilities for public
safety has already been identified. In the Report on Software Defined Radio Technology for Public
Safety 2 , and based on subsequent work, the following potential observations were made on the role
of cognitive radio capabilities for public safety:
1. The first responder can better focus on the incident/threat by eliminating radio operations
ranging from routine to complex through the use of cognitive radio applications to:
a. Be aware of its RF environment (e.g., vicinity of public safety incident);
b. Detect available and authorized RF resources;
c. Decide how to best operate within the existing infrastructure/network;
d. Use geolocation, spectrum, and network awareness to minimize interference;
e. Automatically reconfigure and connect; and
f. Learn how to perform these steps better the next time.
2. Cognitive radios offer a broad range of RF techniques to improve performance, interoperability,
and efficiency.
3. Cognitive radio is becoming a significant concept for all future communications systems and
devices for two fundamental reasons:
a.
It enhances spectrum efficiency and improves access by making dynamic channel
assignments, taking specialized measures to avoid harmful interference to others, and
reducing unused channel seconds.
b. It enables “intelligent” self-configuring, auto-adapting systems and devices that can handle
the growth trend of complex waveforms and user requirements.
4. Public Safety must carefully balance spectrum efficiency benefits against the critical need for
system reliability, robustness, security, “instant on,” and other application-specific requirements
of the first responder.
This report documents the initial efforts of the Public Safety SIG to follow up their report by
exploring and analyzing the potential value of cognitive applications for public safety.
2.1. Cognitive Radio Technology
Cognitive radio technology (note definitions in the box that follows) is rapidly evolving as a
significant driver of capabilities in new radio systems. Initial capabilities to adapt frequencies
automatically to prevent interference to legacy radio systems was successfully field-demonstrated
under the Defense Advanced Research Projects Agency (DARPA) XG radio program in August,
Demonstrations at the 2007 DySPAN Conference included real-time spectrum
2006.3
sensing/monitoring, secondary spectrum use by cooperating cognitive radios, policy engines, and
2
Ibid., Section 4.3 Role of Cognitive Applications.
Seeling, Frederick W., A Description of the August 2006 XG Demonstrations at Fort A.P. Hill, Proceedings of the
IEEE Conference on Dynamic Spectrum Access Networks (DySPAN), April, 2007.
3
-3-
SDRF
SDRF-07-P-0019-V1.0.0
cognitive radio development platforms.4 Such demonstrations indicate that these basic capabilities
are achievable in the near term. We recognize that from these building blocks it will still be
necessary to develop functional capabilities that not only work with existing radio systems but that
are also proven under field conditions before such capabilities can be adopted for public safety use.
Definitions
(from SDRF Cognitive Radio Definitions. SDR Forum Report SDRF-06-R-0011)
•
Radio: (a) Technology for wirelessly transmitting or receiving electromagnetic radiation to facilitate
transfer of information. (b) System or device incorporating technology as defined in (a). (c) A general term
applied to the use of radio waves—from ITU-R Radio Regulations, Article 1 (Terms and Definitions,
Section 1.4).
•
Software Defined Radio: Radio in which some or all of the physical layer functions are Software Defined.
•
Software Defined: Software defined refers to the use of software processing within the radio system or
device to implement operating (but not control) functions.
•
Adaptive Radio: Radio in which communications systems have a means of monitoring their own
performance and a means of varying their own parameters by closed-loop action to improve their
performance.
•
Cognitive Radio:
o a) Radio in which communication systems are aware of their environment and internal state and can
make decisions about their radio operating behavior based on that information. The environmental
information may or may not include location information related to communication systems.
o b) Radio (as defined in a.) that utilizes Software Defined Radio, Adaptive Radio, and other technologies
to automatically adjust its behavior or operations to achieve desired objectives
•
Cognitive Network: A cognitive network is a network able to establish links between its Cognitive Radio
Nodes to provide connectivity, and to adjust its connectivity to adapt to changes in topology, operating
conditions, or user needs. A cognitive network consists of nodes that are cognitive radios. In such a
network, the cognitive abilities of the radio nodes include awareness of the network environment, network
state and topology, and shared awareness obtained by exchanging information with neighboring nodes or
other network accessible information sources. Cognitive decision making considers this collective
information and is performed in coordination and/or collaboration with other nodes.
•
Public safety: the function of safeguarding the lives and property of the general population.
•
First responder: an individual from a police department, fire department, emergency medical team, or
other similar organization. His/her responsibilities when responding to an incident are to take necessary
action to save lives, protect the welfare of others, and inform other personnel of any potential danger at the
scene of an incident.
Often the terms “first responder,” “emergency services,” and “public safety” are used interchangeably. These
terms generally refer to the same group of people and functions. We use the term “public safety” in this report,
but in the International Telecommunication Union (ITU) and in many parts of the world, the phrase “public
protection and disaster relief (PPDR)” is the agreed terminology. For convenience, we have used the term public
safety consistently throughout the report, but the acronym “PPDR” could be substituted in all occurrences
without changing the meaning of the text or the objectives of the report.
4
A summary of each of the demonstrated capabilities is available at
http://www.ieee-dyspan.org/Demonstrations.html.
-4-
SDRF
SDRF-07-P-0019-V1.0.0
We also recognize that there are major issues that must be addressed to realize the potential of
cognitive radios. One of the challenges in dynamic spectrum access is the hidden node problem—
assuming that a frequency can be utilized when in fact it is already in use by a transmitter or
receiver “hidden” (electromagnetically) from the cognitive radio. Another important issue is that
cognitive capabilities assume some level of reconfigurability of the radio which could have
implications on the size, weight, and/or power requirements of a portable public safety radio. While
not discounting the significance of these challenges, progress to date in this field suggests that they
can be resolved to a level sufficient to realize the use cases outlined in this document.
2.2. Assumptions
The analysis of the scenarios in the remainder of this document, and the conclusions that are
drawn in Section 6, are based on the following assumptions.
1. The focus of this document is on derivation of functional capabilities from identified use
cases for enhanced communications capability. We recognize that the technology to realize
these use cases is generally not available in current public safety radio systems. In fact, the
capabilities envisioned in this document range from those that exist in some types of radios
(but have not generally been implemented in public safety radios) to other capabilities that
may require additional research & development. While attempting to be forward-looking
we also limited the scope of capabilities to those that could be reasonably achieved with
extensions of technology that is at least in the research stage.
2. As noted above, the capabilities defined in the use cases include technologies that have yet
to be fully developed, and as such, the cost of implementing proposed capabilities is not
addressed explicitly. The Public Safety SIG is concurrently developing cost models for
analyzing the cost-benefit tradeoffs of proposed SDR and cognitive capabilities. These cost
models, upon completion, can be applied to the functional capabilities identified in this
document to support analysis of the cost implications and tradeoffs associated with
implementation of the identified capabilities.
3. The proposed use cases are not limited by existing regulatory regimes. We have attempted
to be realistic in what regulatory changes are feasible. But also, given a compelling use case
for public safety, we assume that the regulatory community would consider appropriate
changes to existing rules, so “feasibility” is not defined in terms of current thinking but
rather in terms of the use cases defined in this document. Thus for each use case
documented in this report, we also identify regulatory issues that may need to be addressed
to enhance the ability of first responders to communicate more effectively than is currently
available. We also note that regulatory perspectives differ by country and world region, and
adoption of regulatory changes identified in this document will vary.
4. In general we use the term “cognitive capabilities” in this document to reinforce the concept
that the cognition required to support public safety communications is not likely to fully
reside in a single radio or device. More likely, the combination of capabilities to be aware
of the environment, make decisions about how to enhance the performance of the
communications capability, and the reconfiguration of the infrastructure and subscriber
equipment to achieve performance enhancement, will be distributed in multiple nodes
within the public safety communications systems.
-5-
SDRF
SDRF-07-P-0019-V1.0.0
5. Many of the circumstances that impacted response in the London bombing scenario have
already been addressed through deployment of additional communications capabilities that
do not involve cognitive capabilities. If all events are known a priori, non-cognitive
solutions can be implemented to account for those events. However, major incidents and
disasters are often characterized by circumstances that are beyond the scope of planning; in
addition, fiscal realities preclude implementing contingencies for all possible situations. The
real power of cognitive capabilities is to rapidly adjust to changes in the operating
environment in order to maintain communications in the face of dynamic and often
unanticipated circumstances. Thus we assume that the use cases discussed in this document
are not specific cases that are optimally addressed by cognitive capabilities, but rather
representative of a class of cases that can be effectively addressed by cognitive capabilities.
6. Communications in incident response are based on guidelines of appropriate multi-agency
incident command and coordination, such as the Gold and Silver coordinating groups in
London, and comparable incident command structures elsewhere, such as the National
Incident Management System (NIMS) in the United States.
7. Scenarios assume non-cognitive legacy radios will continue to be used, and need to be
considered in the solutions provided in the use cases. We assume that any implementation of
cognitive radio capabilities will include capability to interoperate with legacy (noncognitive) radios.
8. There may be the need to connect the public safety radio system with non-public safety
radios such as commercial cellular networks. Part of the use case analysis addresses the
technical capabilities required to ensure that the public safety radios/networks are not
compromised by the inclusion of the non-public safety radios.
-6-
SDRF
SDRF-07-P-0019-V1.0.0
3. METHODOLOGY
The overall methodology is intended to establish use cases within the context of multiple
scenarios based on actual or credible events relevant to public safety. Each scenario will be
analyzed in detail to determine how cognitive radio capabilities could positively impact the
communications of the public safety activity. The results of each individual scenario analysis will
then be compiled to create a final analysis of the potential application of cognitive radios for public
safety applications, along with the technical, regulatory, and procedural issues that must be
considered and addressed to realize the enhanced communications capabilities.
In each scenario we analyze the events to derive use cases. This document includes analysis of
the 7 July bombings in London. Subsequent reports in this series of documents will include analysis
of other scenarios. The analysis of each scenario will follow a common approach:
•
Develop a timeline of events in the scenario.
•
Identify points in the timeline in which cognitive capabilities could enhance
communications (the cognitive use cases).
•
Analyze these use cases in terms of the technical, regulatory, and procedural issues
that need to be addressed to achieve the enhanced communications.
•
Review the use cases with public safety practitioners to ensure relevance and validity.
The remainder of this section describes this approach in more detail.
3.1. Select Scenarios
The first step in the overall methodology is to identify scenarios for analysis. The scenarios
must be sufficiently rich in activity to support analysis of a broad segment of use cases, with
sufficient realism to maintain credibility with the stakeholders (in the case of the public safety
community), and sufficient detail to allow derivation of relevant cognitive radio capabilities from
the use cases. The initial scenario to be selected was the series of bombings that took place in
London on July 7, 2005. It was selected based on the credibility of a real scenario and the extensive
documentation in after action reports5 . This scenario provided an excellent starting point, and
highlighted certain areas for cognitive capabilities, but did not reflect the whole breadth of possible
use cases. The additional scenarios may be either actual events or hypothetical scenarios developed
for analysis and training purposes, and will be analyzed in a similar manner.
3.2. Develop Timeline
The second step in the analysis is to establish a detailed timeline of scenario events. In the case
of the London bombing scenario, the timeline was in place and detailed in the after action reports.
In some places in the timeline we supplemented the existing documentation by postulating
intermediate events or events that were implied but not articulated in the after action documentation
as communications events.
5
Greater London Authority, Report of the 7 July Review Committee, June 2006.
-7-
SDRF
SDRF-07-P-0019-V1.0.0
3.3. Identify Use Case
Having established the timeline, the next step is to identify those situations in which cognitive
capabilities and dynamic spectrum access could have positively impacted the response(s) defined in
the scenario. In analyzing the timelines, we consider the following types of questions to determine
whether cognitive radios could enhance communications capabilities:
•
Are there communications capabilities and links that are needed but do not exist?
•
Are there communications capabilities and links that are established but need to be
established in a more timely manner?
•
Are there communications capabilities and links that are established in a timely
manner whose performance needs to be improved?
By posing these questions in comparison with the timeline of events, we identify situations in which
cognitive capabilities could enhance communications capabilities. The applications of cognitive
capabilities in these situations are defined as cognitive use cases.
3.4. Analyze Use Case
For each identified use case, the next step is to analyze that use case with respect to:
•
how it would impact the communications capability of the responders,
•
how it would potentially impact the response scenario, and
•
any potential negative impacts that could occur if the cognitive capability were part of
the scenario.
We recognize that the use cases described in Section 4 require functional capabilities that are not
necessarily available with current technology. (In fact, one of the purposes of this document is to
provide input to the SDR Forum’s Cognitive Radio Working Group to conduct a gap analysis to
identify technology readiness needs.) As the focus of this document is on functional capabilities
rather than available technologies, we deliberately attempted to place as few constraints as possible
on technologies that would need to be deployed to realize the use cases. We considered capabilities
that were realistic even if they would require future research and development to be implemented.
However, the approach taken was to lay a foundation of capabilities and allow other researchers to
assess the level of technological advancement required to realize the proposed capability.
Likewise, regulatory, policy, and procedural considerations did not constrain the identification
of potential use cases. Regulatory and procedural implications and/or changes were noted in the
analysis section.
In considering the use cases, we identify five aspects of the communications environment that
are useful in framing key issues and challenges. Each of them may change over time.
1. Physical: This aspect is concerned with the physical world, including issues of geography,
geometry, topography, proximity, density, RF propagation characteristics, and locale. What
resources are where, and how can they get to where they are needed? How large is the area
for which communications coverage is required? What is the geometric distribution of
injured people, hospitals with capacity available, and transport to move them? How are
-8-
SDRF
SDRF-07-P-0019-V1.0.0
responders moving over time and how is the geographic layout of the responders changing
over time (e.g., expanding a perimeter)? What are the facilities for recharging radios?
2. Network: This item deals with the technical issues of how information flows, both
normally and in response to the emergency, and how failed systems are restored. How are
radio systems structured, and how do they connect with other networks such as personal
area networks, commercial systems such as the telephone network or WiFi/WiMAX
capabilities? Is there a need for interoperability? How much bandwidth is needed? What
Quality of Service is needed? What kinds of terminals are available? How do all the
different agencies talk with each other and distribute data? Does the network use
infrastructure, repeaters, direct peer to peer, or an ad hoc mesh? What authentication
mechanisms are used in the network?
What cryptographic algorithms support
communications security? How are keys distributed? Can radio functionality be modified
over the network, or is hands-on intervention needed?
3. Procedural: This issue deals with the role of people in the system, including authority,
command, control, operating procedures, communications security procedures, and
activation of contingency plans. Who develops contingency plans? Are there memoranda
of understanding (MoU) in place to establish communications interoperability and ensure a
chain of command is in place at the time of the incident? Who has command and control of
the situation? Who authorizes individuals or groups to operate radios? What is the
registration process for new communications devices requiring authentication? Who
reviews requests for registration? How are cryptographic keys managed and what are the
respective roles of humans and automation in key management? Who reprograms radios?
Who assigns tasks and/or roles to individual units?
4. Regulatory: Regulators administer the use of spectrum, issue licenses for radio operation,
resolve issues of interference, and make rules for radio operation. What modulation
techniques and frequencies can be used? What are the rules for operation in unlicensed
spectrum? How can extra spectrum be made available under emergency conditions? Are
spectrum sharing agreements in place? What agencies have jurisdiction?
5. Chronological: Time is an overarching consideration that applies to all of the other aspects
listed above. During the time before a specific event, there is time for establishing
organizations, procuring equipment, recruiting and training personnel, building networks,
defining policies and procedures, and development of contingency plans. When an event
occurs, the first problem is awareness that something has happened, and learning enough to
assess the situation. Then decisions are needed to determine the nature of the response,
what resources to commit to it, and what actions to pursue. Operations continue until the
emergency is resolved, and then the units involved stand down. After the event is over, an
after-action review considers how well the operations were executed, and what can be done
to improve preparedness for subsequent emergencies. A primary value of cognitive radio
capabilities is to adapt to the changing environment much more rapidly than present
systems.
-9-
SDRF
SDRF-07-P-0019-V1.0.0
3.5. Review Results with Public Safety
To ensure that the assumptions made concerning public safety incident response were valid, the
analysis was reviewed by representatives of the public safety community. Comments were solicited
and received from members of the National Institute of Justice (NIJ) Communications Technology
Working Group and representatives of the ITS UK who also have experience in law enforcement
and homeland security. Their comments were provided as individuals and not as official positions
of their organizations, and were valuable contributions to the formation of this document. Their
inputs provided valuable guidance and credibility for the analysis.
-10-
SDRF
SDRF-07-P-0019-V1.0.0
4. LONDON BOMBINGS OF 7 JULY 2005
“For hundreds of thousands of people commuting into London, the morning of July 7, 2005
began just like any other. But at the peak of the rush hour, bombs were detonated in three crowded
subway trains and aboard a London bus. At least 52 people died, along with four bombers, and 700
were injured.”6 As shown in Figure 4-1, the explosions were spread out across several incident
sites, and as shown in the timeline in Appendix A, occurred over the course of an hour.
This event provides a real-world scenario that illustrates the significant challenges in responding
to a terrorist event. As noted in the Report of the 7 July Review Committee, 7 the response was
characterized by countless acts of heroism and compassion by first responders and ordinary citizens
alike, and that both the communications and the command and control capabilities generally
functioned as well as could be expected under the circumstances. 8 The purpose of this section is to
use the sequence of events that occurred on that day and the communications that occurred as part
of the response to identify circumstances in which future cognitive capabilities could provide more
efficient and effective communications in similar situations.
Figure 4-1. Map of the London Bombing Incident Locations
4.1. Scenario Assumptions
Since this scenario is drawn from actual events, our only assumption is that the events unfolded
as reconstructed in the after-action reports. In some cases we also assume the existence of
communications not specifically identified in the after action reports but logically based on incident
6
CNN, “Bombers Target London,” available at http://www.cnn.com/SPECIALS/2005/london.bombing/.
Greater London Authority, Report of the 7 July Review Committee, June 2006.
8
London Emergency Services Liaison Panel, “Major Incident Procedure Manual”, 6th edition, July 2004.
7
-11-
SDRF
SDRF-07-P-0019-V1.0.0
response and subsequent events (e.g., we assume units were dispatched to a scene if they arrived at
the scene but the dispatch activity is not specifically noted).
4.2. Scenario Timeline
The timeline of the London bombing incident was derived from the after action reports and
included as Appendix A. In the scenario there are a number of places in the timeline (use cases)
where the Public Safety SIG identified potential applications for cognitive capabilities to improve
the communications capability that existed at that time. In the timeline these events are shown with
a yellow shaded cell in the table. Section 4.3 provides a detailed discussion of these use cases.
4.3. Cognitive Use Case Discussion
A number of potential cognitive use cases have been identified based on the scenario timeline as
described in the preceding section. In this section, each use case is discussed in much greater detail.
The use cases are ordered in descending priority based on operational relevance and feasibility as
provided by the public safety practitioners who provided input to and feedback on the report.
4.3.1. Use Case 1: Network Extension for Coverage and Reachback
Cognitive radio capabilities could be used to automatically reconfigure radios to include a
repeater capability to extend network coverage to areas where radios are otherwise cut off from their
infrastructure, particularly during initial response to an incident prior to additional communications
resources being deployed.
4.3.1.1 Summary of Scenario Situation
In terms of the aspects of the public safety communications environment:
1. Physical: Bombs exploded on three London Underground trains inside tunnels with varying
distances to the nearest station. Some passengers were severely wounded. There was no
light. The only escape was by walking through the tunnel to the nearest station. Responders
had to walk to the scene through the tunnel.
A number of different agencies, including Metropolitan Police, British Transport Police, the
London Fire Brigade, and London Ambulance responded to the emergency by entering
tunnels through the nearest station.
2. Network: Once police and fire responders went into the tunnels, their radios lost
connectivity to the above-ground infrastructure. The only means for responders to
communicate back to their respective command centers and any above ground personnel
was to walk to the nearest station and position themselves at the entrance to the Metro
system. Individual radios were not capable of exploiting peer-to-peer capabilities to provide
network extension to connect isolated nodes to the network.
3. Procedural: Responders had adequate authority to communicate on their own networks, but
as noted under Network were unable to do so. Procedures were established to maintain
some flow of information by having responders communicate to (above ground) command
-12-
SDRF
SDRF-07-P-0019-V1.0.0
centers from the entrance to the tunnels, but that process required responders to walk from
the scene to the entrance, which took as long as 15 minutes in some cases.
4. Regulatory: No regulatory issues were involved as the situation reflected an inability to
communicate on licensed frequencies due to physical constraints.
5. Chronological: Prior planning had been performed, but infrastructure damage precluded the
use of some communications capabilities that were in place and part of the plan.
The other critical chronological consideration is the amount of time (as much as 15 minutes)
required to move information from the scene of the accident underground to the command centers.
4.3.1.2 Description of Use Case
Cognitive radio technology could be implemented to reconfigure responders’ radios to create an
extension to the existing network. This network extension would allow transmissions to be passed
back and forth from the incident site along a network of individual responder radios operating in
peer-to-peer mode to a radio which can communicate with the main radio system/network. A radio
would be positioned where it could maintain connectivity with the above-ground infrastructure
(such as at an opening to the tunnel) and function as a repeater to bridge between the otherwise
disconnected radios and the infrastructure. Depending on distribution of radios in the tunnels,
additional radios could also be automatically reconfigured to act as repeaters among the
disconnected radios.
The concept is illustrated in Figure 4-2. As shown, communications is enabled between
personnel at the opening of the tunnel to dispatch and emergency management centers, but not from
responders at the scene of the explosion in the tunnels. The concept of the network extension
capability is reflected in the additional links that could be established automatically among
responders otherwise cut off from communication with the above-ground system. This provides
immediate restoration of communications for all users without requiring additional equipment at the
scene.
-13-
SDRF
SDRF-07-P-0019-V1.0.0
Street level Officer
Dispatch/Emergency
Operations
Above Ground
Below Ground
No communications
directly from subway
to dispatch
Links enabled
automatically
based on cognitive
capability
Intermediate Radio
(automatically
reconfigured as repeater)
On scene officer
Figure 4-2. Network Coverage Extension Use Case Example
The impact of these capabilities is that on-scene responders would have direct communications
to command centers without leaving the incident scene or resorting to runners that delayed
communications by as much as 15 minutes. With respect to the specific aspects of the scenario
situation noted in Section 4.3.1.1, this use case would result in the following:
1. Physical: No change from the physical situation described above.
2. Network: Responders would maintain connectivity with their network at all times
regardless of where they were located. (Note that it would be possible to achieve the same
effect by deploying repeaters at strategic locations to create the necessary extension.
However, the cognitive capability has significant advantages that justify this use case:
a. Cognitive capabilities and the ability to reconfigure radios would provide the network
extension immediately, rather than after the period of time necessary to deploy repeaters.
-14-
SDRF
SDRF-07-P-0019-V1.0.0
b. Cognitive capabilities would automatically determine appropriate network
configuration. Repeaters would require manual determination of the repeater location,
frequencies, and so on.
c. Cognitive capabilities would allow the network extension to accommodate the dynamics
of the response, as users arrive and leave, as the physical location of the responders
changes, and so on.
3. Procedural: In general the procedures normally used for communications among
responders and between the responders and the above ground command centers would be
followed in this situation. Additional procedures would be followed to position radios as
needed to function as repeaters to ensure connectivity. However, unless the radios acting as
repeaters have a full duplex capability, the responders must allow for a variable time lag for
their message to be received and stored and then re-transmitted at each repeater and for an
appropriate response back from the other party(ies) involved in the communications.
4. Regulatory: No significant regulatory change would be required as this capability assumes
operation on existing licensed frequencies, however regulatory approval of operating mode
changes would be required. See Section 4.3.1.4.
5. Chronological: The pre-planning would still be effective as network connectivity would be
maintained despite infrastructure damage or the response by responders whose radio system
was not normally extended into the tunnels. In addition, the timelines to move information
from the scene to the command centers could be reduced from as much 15 minutes to a few
seconds.
4.3.1.3 Functional Capabilities
There are a number of functional capabilities assumed by this cognitive use case. First, the
radios must be capable of being reconfigured to function as a network extension (e.g., the radios can
operate on appropriate spectrum; the radios have reconfiguration algorithms, and so on). Second,
there must be some level of cognitive capability for a collection of radios to “understand” that they
have lost their ability to communicate with the system infrastructure. More specifically, radios must
be capable of:
•
Determining that they are disconnected from the system infrastructure;
•
Finding and identifying peer radios;
•
Identifying and authenticating reconfigurable compatible radios;
•
Determining which radio/radios is/are within coverage of the infrastructure and which radios
are not within the coverage of the infrastructure;
•
Forming a satisfactory network extension route to the infrastructure from each affected
radio using non-interfering frequencies for each “hop”;
•
Adjusting the network as responders arrive and depart from the area where coverage is
unavailable; and
•
Preserving the level of security of the baseline network in the network extensions.
•
Providing either full duplex (simultaneous receive and transmit) operation or including a
“store and forward” capability for user voice and/or data communications.
-15-
SDRF
SDRF-07-P-0019-V1.0.0
There are a number of approaches that could be utilized to achieve the network extension, such
as ad hoc or mesh networks. The feasibility of existing protocols to accomplish this is a relevant
research topic.
4.3.1.4 Regulatory Implications
The proposed capability would rely on existing peer-to-peer modes of operation for which
regulatory rules are already in place. Thus no major regulatory changes are required (although air
interface could change from being used in simplex mode to semi-duplex or duplex mode.)
However, it is possible that the spectrum and protocols required to implement such a capability
would require the ability to utilize spectrum not routinely licensed to such users under current
regulatory rules. Thus it may be necessary to consider allowing use of spectrum for such purposes.
Part of the cognitive capability of identifying the disconnection from system infrastructure could be
to ensure that spectrum used for peer-to-peer communication would not interfere with other users.
In the case of the tunnel bombing, such interference would be unlikely because the area in question
is generally cut off from most above-ground infrastructure. However, to generalize this use case to
other situations in addition to the tunnel scenario, channel assignments should be made on approved
frequencies.
4.3.1.5 Policy Implications
As noted above, the objective of this use case is to seamlessly restore lost connectivity of
responders who have already been authorized and authenticated to use the network. Ideally
responders would be unaware of the reconfiguration of radios, although an audio or visual alert
should be included to indicate that the radio is following the network extension protocols. Policies
governing use of the network (e.g., who may use the network, radio protocols, use of channel)
would be the same for the network extension as when the radio is within range of the infrastructure.
Some procedures will need to be modified or added to reflect that some radio behavior, such as
the rebroadcast of voice transmissions, may be evident to the responders because of response
delays. Training will be generated to be appropriate so that users will be familiar with the
differences in radio performance and behavior. In addition, training will be established for
responders to understand the impact of physical location and how best to deploy, particularly if they
were going to carry a radio that can be reconfigured as a repeater. Consideration should also be
given to allowing individual first responders the option of excluding their radio from participating in
the network (see discussion in Section 5.2).
Network management policies (e.g., machine readable policies) that govern the manner in
which radios are reconfigured to achieve network extension will need to be implemented in advance
to facilitate such operations. However the impact on usage behavior should be minimal.
Some training will also be needed to fully exploit such capabilities. For example, responders
would need to understand how the radios respond when outside network coverage. Users should be
familiar with performance aspects of the radios such as changes in delay times, capabilities and
performance of the radio when acting as a repeater, the impact of physical location of the radios on
network performance, and similar information.
Policies and procedures will need to be established for operation of radios that end up being
used as repeaters. For example, a responder whose radio is used as a repeater may need to stay in a
specific physical location and “man” the position to maintain the communication link, and may not
be able to participate in other functions of the response. Agencies will need to consider the tradeoffs
-16-
SDRF
SDRF-07-P-0019-V1.0.0
of how to allocate responders accordingly. Procedures may also be established for responders
“dropping off” radios as needed to allow the network extension to function (here again the issue is
how best to allocate resources, with the radio and the responder being considered separate
resources). Given the proposed capabilities of the radios, spare or cache radios which can be rapidly
delivered to the scene may also be part of an incident response strategy.
4.3.2. Use Case 2: Dynamically Access Additional Spectrum
At several points in the scenario there were communications difficulties because of the sheer
volume of calls on the voice communications networks. Dynamic spectrum access, or the ability
for cognitive radios to identify unused or underutilized spectrum, could be a solution in this scenario
and provide a means for expanding capacity when needed.
We understand that there are significant technical issues associated with implementation of this
use case that are dependent on the technology used in the system infrastructure. The focus of this
discussion is the overall desirability and benefit to public safety of being able to expand capacity in
emergency situations in a timely manner.
Note, however, that most dynamic spectrum access approaches assume that the user is operating
as a secondary user, and able to relocate to other spectrum as needed if a primary user utilizes the
spectrum. This use case in this scenario would be significantly different—in an emergency
situation, dynamic spectrum access for responders would be as a primary user—to specifically
appropriate spectrum that is not being used, or can be appropriated for emergency use.
4.3.2.1 Summary of Scenario Situation
In terms of the aspects of public safety communications:
1. Physical: In the process of responding to the situation, the density of radios and access
attempts overloaded some infrastructure elements in a specific geographic area. Key
considerations include that the scenario took place in a densely populated urban area during
a workday. In addition, multiple incident locations created demand for services that
impacted the system as a whole.
2. Network: Based on the demand exceeding capacity, access control mechanisms
(ACCOLC) were invoked in the area around Aldgate East to block access for some users,
including first responders that did not have priority access.
3. Procedural: A significant decision process was executed to determine whether or not to
invoke Access Overload Control (ACCOLC) 9 when the mobile phone system could not
handle the number of attempted calls. Part of the decision process was assessment of the
impact of ACCOLC on responders who were supporting the response but whose radios did
not have priority.
9
We recognize that the increasing deployment of Airwave (a dedicated public safety communications system) into
the police and fire services is reducing the use of public networks by first responders, and therefore reducing the
chances of a recurrence of the specific circumstances under which ACCOLC was invoked in this scenario.
Nevertheless, this scenario highlights the general challenge of obtaining adequate capacity for first responder
communications in an escalating event, for which dynamic spectrum access is an important use case for cognitive
radio capabilities.
-17-
SDRF
SDRF-07-P-0019-V1.0.0
4. Regulatory: No regulatory procedures existed for dynamic allocation and use of spectrum
outside the previously licensed frequencies.
5. Chronological: The ACCOLC decision process occurred as the response to the bombing
events was unfolding.
4.3.2.2 Description of the Use Case
This use case involves identifying and utilizing spectrum not normally utilized by the system—
in this case the mobile phone system.
There are three different approaches that can be considered to realize this use case, as outlined
(underlined) below:
Pre-defined agreement among organizations: One approach to dynamic spectrum access, taking
advantage of reconfigurable radios/cell phones, is to establish agreements among organizations that
would allow a non-licensed authorized user to utilize additional spectrum under defined
circumstances and by mutual agreement. Implementation of this cognitive capability may be
limited to the ability to identify the channel loading limits that would require accessing additional
spectrum. The cognitive capability may also be used to manage network and subscriber
reconfiguration to enhance the utilization of the allocated spectrum. There is a broad range of
potential types of agreements under which spectrum could be dynamically accessed. The following
is by no means conclusive but serves to provide a range of possibilities:
•
Dynamic spectrum access that is triggered by a pre-defined event, such as reaching a
capacity limit;
•
Dynamic spectrum access that occurs when one organization requests access and the
licensed organization grants it (e.g., spectrum mutual aid).
•
Dynamic spectrum access granted to another user (spectrum leasing) or to a
secondary user on a non-interfering basis.
Emergency declaration: Another approach to dynamic spectrum access, again requiring the
ability to reconfigure radios/cell phones is to establish rules by which some spectrum (licensed for
other services) is accessed for emergency response under a governmental declaration. Here again
the cognitive capability may be limited to identifying the load circumstances under which access of
additional spectrum is appropriate, or may be used to manage network and subscriber
reconfiguration to enhance spectrum utilization.
Identify unused or underutilized spectrum not licensed to the network: Another approach to
dynamic spectrum access is to monitor spectrum utilization in frequencies not licensed to the
network, identify spectrum which is unused or underutilized (“white space”), and reconfigure the
network and subscriber equipment to utilize that spectrum. Clearly this type of dynamic spectrum
access would be limited to emergency situations and only be allowed under clearly defined
circumstances (such as a governmental declaration). Cognitive capabilities would be required to
identify available spectrum and to reconfigure the network and subscribers accordingly.
1. Physical: No change from the physical situation described above.
2. Network: In this scenario, the network congestion would be relieved by providing more
spectrum for use by all network users including first responders. If the infrastructure has a
cellular architecture, it may be possible to dynamically reallocate the channel distribution to
-18-
SDRF
SDRF-07-P-0019-V1.0.0
create additional capacity in the afflicted cells. Alternately, there may be a place in nearby
spectrum where some other service can be pre-empted to satisfy the demand.
3. Procedural: There are a number of procedural decisions involved in implementing this use
case. Key procedures include determining at what point to invoke dynamic spectrum access
procedures, procedures for identifying spectrum that can be utilized, and when to release the
bandwidth. The specific procedural implications are a function of the implementation
approach. For example, this use case could be based on a fully automated determination of
the need for additional spectrum and the spectrum to be utilized; in other implementations
there may be a human in the decision loop, in which case the procedures for making such a
decision must be defined. Also, different procedures may be appropriate depending on
whether the additional spectrum is based on a pre-defined procedure or agreement, or
whether additional spectrum is identified in real-time during the course of an incident.
4. Regulatory: The regulatory implications of dynamically accessing spectrum depend on the
approach (as outlined above) that is used. Some pre-defined agreements among
organizations may be feasible under existing regulatory rules, particularly if the spectrum is
allocated under the same service rules or if the rules explicitly provide for secondary
spectrum usage. Rules may require modification if the dynamically allocated spectrum is
normally allocated under different service rules, or if there is no explicit allowance for such
agreements to be put in place. In general the approach that allows spectrum to be accessed
for emergency response is not embodied in existing regulations and would need to be added
to allow this approach. Likewise, the ability to identify and access unused or underutilized
spectrum not licensed to the network is not generally part of existing regulations. We
recognize that these regulatory changes can involve sweeping changes to how spectrum is
utilized in emergency situations, and that crafting rules which balance the needs of
emergency response and other legitimate uses of spectrum during emergencies will require
extensive research, development, and public discussion.
5. Chronological: This kind of Spectrum Sharing would require a significant amount of
advanced detailed planning. Plans can have varying degrees of dynamic range. Switching
from one fixed plan to another is easier than dynamic cognitive problem solving in real time,
but more likely would result in less efficient spectrum utilization. Also note that
relinquishing spectrum that has been utilized to facilitate emergency response must be done
in a timely manner to have value greater than present systems.
4.3.2.3 Functional Capabilities
Dynamic spectrum access implies a number of functional capabilities, as described below.
•
The network must be able to identify capacity loading that meets whatever criteria are
in place to initiate the dynamic spectrum access.
•
The network must be capable of identifying spectrum resources that can be utilized to
offload some calls. There are two possible approaches to identifying additional
spectrum.
o First, there may be established agreements in place that under certain
circumstances spectrum normally used for one purpose is made available to
support communications networks being utilized in an emergency. Such
identification could be based on established agreements among spectrum
-19-
SDRF
SDRF-07-P-0019-V1.0.0
“owners” or based on allocation of spectrum for emergency use in the event of
a certain level of emergency.
o Alternatively, cognitive capabilities to search for underutilized spectrum
(“white space”) that could be dynamically accessed. Note that in this case a
scheme must be implemented to manage the hidden node problem. Also, the
network must be able to support the ability to deconflict the situation if
multiple users attempt to access the same available spectrum “white space”.
•
The network infrastructure must be able to reconfigure to use the new spectrum. If
the system is a trunked system, the network must be able to incorporate additional
frequency options into the system. Network transmitters and receivers must be able
to be reconfigured to utilize the additional spectrum. If the additional spectrum is
based on a pre-defined agreement, frequencies may be pre-programmed, in which
case only an execution command is required to access the additional spectrum.
•
Subscriber equipment must be able to reconfigure to use the new spectrum, i.e., must
be able to transmit and receive on the additional frequencies.
•
Reconfiguration information must be communicated among the radios and the
network infrastructure to coordinate the utilization of additional spectrum.
•
Dynamic access of spectrum must be consistent with the regulatory requirements of
that spectrum (e.g., in terms of bandwidth, out of band emissions, power
management, location based rules) to ensure that other users in that service are not
adversely impacted by use of a specific frequency.
4.3.2.4 Regulatory Implications
The regulatory implications of this use case depend largely on the manner in which spectrum is
dynamically accessed.
Approaches based on pre-defined agreements among organizations that allow users to utilize
spectrum in emergency situations may require regulatory approval to allow secondary use
(secondary markets, leased spectrum, etc.) of spectrum by non-licensed users.
One potential regulatory change is to allocate spectrum for first responder use that is otherwise
allocated for other non-public safety use during normal conditions (e.g., executive declaration of an
emergency automatically dynamically allocates certain commercial use spectrum for emergency
responder utilization).
Use of licensed spectrum without pre-arrangements is generally not allowed and would require
changes to existing regulations. Use of unlicensed spectrum is generally allowable, although
regulatory changes could recognize public safety priority use of unlicensed spectrum in emergency
situations (just as drivers yield the right of way to emergency vehicles with lights and sirens).
4.3.2.5 Policy Implications
There are a number of policy implications for this use case. Key questions include:
•
What are the circumstances under which spectrum can be allocated as described above?
•
Who has the authority under which a decision to utilize non-licensed spectrum is made?
•
What is the interaction of priority services and dynamically allocated spectrum?
-20-
SDRF
SDRF-07-P-0019-V1.0.0
•
When and how is dynamically allocated spectrum released?
4.3.3. Use Case 3: Temporarily Reconfigure First Responder Communication Device
Priorities
Cognitive radios (in this case, referring to cell phones) might be able to be temporarily
reconfigured with higher priorities based on the circumstances of the emergency responder.
4.3.3.1 Summary of Scenario Situation
In terms of the aspects of public safety communications:
1. Physical: In the process of responding to the situation, the density of radios and access
attempts overloaded some infrastructure elements in a specific geographic area.
2. Network: Mobile phone network resources were being utilized by public users as well as
first responders.
3. Procedural: A significant decision process was executed to determine whether or not to
invoke Access Overload Control (ACCOLC) when the mobile phone system could not
handle the number of attempted calls. Part of the decision process was assessment of the
impact of ACCOLC on responders who were supporting the response but whose radios did
not have priority.
4. Regulatory: No regulatory issues were involved.
5. Chronological: The ACCOLC decision process occurred as the response to the bombing
events was unfolding.
4.3.3.2 Description of the Use Case
The dynamic prioritization use case exploits cognitive capabilities to adjust the priorities of
responders based on the ongoing communications activity as well as the dynamics of incident
response. Priority schemes are implemented in today’s public safety and commercial cellular
systems. The application of cognitive capabilities provides the opportunity to adjust those priorities
to accommodate unanticipated priorities or to manage priority access in real-time.
One of the sources of motivation for this use case comes from one of the major issues that arose
in the London bombing scenario. The high demand for cellular calls motivated the Gold
Coordinating Group to consider activating the Access Overload Control (ACCOLC) to deny access
to the system for any device that did not have the required priority access. One of the
considerations in the decision of the Gold Coordinating Group not to invoke ACCOLC was concern
that the key responders might not be carrying phones that would allow access were ACCOLC to be
invoked.
While the after action reports cited issues surrounding the deliberation to invoke ACCOLC on
the day of the bombing, our analysis led to consideration of another use for cognitive capabilities:
dynamic priorities. In a crisis situation, as demands for system resources rise, it may become
necessary to manage access to the system based on the relative importance of the user and the
communication being transmitted. The concept of the use case is to be able to change those
priorities in real-time as an event unfolds. In the case of emergency responders using a commercial
cell phone network, priority access may be public safety users getting priority access over
-21-
SDRF
SDRF-07-P-0019-V1.0.0
commercial users in the event of emergency situations. For example, in land mobile radio systems,
“man down” alarms get priority over other communications.
While there are differences in typical use of cell phones for incident response between the
United Kingdom and locations in other regions such as the United States, this use case is still
generally applicable. It is not uncommon for first responders and incident command staff to use cell
phones for non-mission critical communications. While not mission-critical, there may still be
significant benefit in managing the priorities of such users. Furthermore, although not the focus of
this particular discussion, the entire concept of dynamic prioritization based on responder role can
be applied to land mobile radio systems as well. Trunking systems today have prioritization
capabilities, but they are statically defined.
The role of cognitive capabilities here is in the ability to adjust in real time those priorities based
on the unfolding events of the incident, communications resources demands and availability, and
the changing roles of individual responders over the course of an event. In the case of the London
bombing scenario, a capability that would have enabled responders who did not have ACCOLCenabled devices (cell phones) would be to have devices reconfigured over-the-air and in real time.
This could have eliminated the risk of responders being denied access to the system in the event that
ACCOLC was invoked. These cognitive capabilities could provide more sophisticated and
dynamic access management for radio/cellular systems.
The scenario situation would be described as follows:
1. Physical: No change from the physical situation described above.
2. Network: First responders would be assigned a priority based on their role in support of the
response. Priority modifications would be downloaded to the first responders’ mobile
phones as needed. In addition, cognitive capabilities in the network management would
recognize the increasing load level and congestion levels and block access to lower priority
calls as needed. User mobile phones would also have a cognitive capability that indicates
that user access has been blocked so that the system loading is not made worse by persistent
access attempts.
3. Procedural: As discussed in Section 4.3.2.3, there are several approaches to deploying this
cognitive use case. Appropriate procedures will be needed, and depend on what particular
approach is followed. For example, if individual responders are allowed to change (or
request to change) their priority, policies and procedures need to be defined to govern the
circumstances and steps to be followed by responders. Likewise, policies and procedures
for any request approvals or assignment of priorities as described in the following section
will be required.
4. Regulatory: Mechanisms for over the air reprovisioning of mobile phones may require
regulatory modifications.
5. Chronological: Policy and procedures would need to be addressed as part of system
planning. During an incident, invocation of access controls would take place upon
activation of a set of trigger conditions. When demand no longer exceeds capacity, then the
access control mechanisms can be removed, although the normal feedback loop stability
criteria must be observed to avoid an on-off-on-off pathology.
-22-
SDRF
SDRF-07-P-0019-V1.0.0
4.3.3.3 Functional Capabilities
A number of capabilities must be available in order to realize this cognitive use case. First,
there must be a mechanism to determine those first responders who have a legitimate need to have
priority access to the communications network. Access to the network itself has already been
established, i.e., the network has already recognized and authenticated the first responder’s cell
phone (responder’s device). The required capability is to establish that the circumstances of that
particular user warrant a level of priority greater than the priority level currently granted to the
responder’s device.
The definition and assignment of priorities can incorporate a number of different elements of
incident response and management. For example, priority assignments could be based on:
•
The roles within the response that have been assigned to the individual responder’s device;
•
Physical location of the responder’s device;
•
Service of the responder’s device (e.g., EMS priority over law enforcement);
•
Type of data being communicated;
•
Role of the user in the communications process.
There is a potentially broad range of complexity and sophistication of the cognitive capabilities
implied by this use case. At the simplest level, assuming that priorities can be dynamically
modified, radios could be reconfigured either by the individual responder or manually by a network
operator without utilizing any cognitive capability. However, manually determining priorities for
individual radios is not very practical for large scale incidents. Relatively simple cognitive
capabilities10 could be implemented to associate priorities with responder assignments, physical
location, and/or service. More sophisticated cognitive capabilities could assign priorities
automatically based on a variety of parameters associated with the communications of the response,
or even in a predictive mode to anticipate, rather than react to, the dynamic needs of the responders.
The advantage of role based priorities (supplemented by other ad hoc assignable methods) is
that preplanning can determine the appropriate priorities for each role in a variety of situations of
varying complexity. Cognitive capabilities might be able to assess the level of complexity involved
and select a suitable priority. All of this is supplemented by the user controlled methods delineated
as follows.
One approach to considering the different functional capabilities for handling priority
assignments is to consider that there are three possibilities for requesting changes in priority—the
responder, some central authority (e.g., incident command, incident communications leader), and
the communication network itself. Note that in the case of the communications network, the actual
functionality could be distributed between the subscriber unit and the network infrastructure, but the
request or the authorization is made automatically without human initiative. Each of these entities
may also authorize the requested priority change. This leads to nine possible approaches to priority
assignment as shown in Table 1.
10
Relatively simple in this context refers to the notion that the complexity of an algorithm to assign higher priorities
to responders in a defined location is low; we recognize that the ability to reconfigure cell phones or other
communications devices dynamically is a challenging issue.
-23-
SDRF
SDRF-07-P-0019-V1.0.0
Table 1. Possible Dynamic Prioritization Approaches
Authorized
by
Individual Responder
Central Authority
Network
Individual
Responder
Priority is controlled by
individual responder
Individual requests are
granted “manually” by
central authority, would
not require cognitive
capabilities.
Cognitive capability to
respond to individual
request.
Central
Authority
Priority changes are
initiated by central
authority and “accepted”
by individual responder.
Central authority makes
unilateral decisions
regarding individual
responder priorities.
Cognitive capabilities in
the communications
network evaluate
requests initiated by
central authority
Network
Cognitive capabilities in
network “recommend”
priority change to
individual responder
who must “accept” the
change.
Cognitive capabilities in
network “recommend”
priority changes to
central authority who
must “accept” the
change.
Fully automated
capability for priority
management with no
human in the decision
loop.
Requested by
While any of the above approaches is possible, we recognize that not all approaches will be
appropriate for all situations, and user requirements for a specific system may well dictate that only
one of the above approaches be implemented in a particular system. In addition, we recognize that a
investment may be required to maintain information on responder credentials and to establish
general policies as well as specific priorities associated with roles assigned to individual responders.
Also note that there are ongoing and planned efforts in developing responder credential
infrastructure to support incident management that can be leveraged to support this use case.
The other significant functional capability is the capability to reconfigure such a responder’s
device. In this situation involving a GSM-based cellular network 11 , when ACCOLC is invoked,
only cell phones with a SIM with priority authorization can access the system; other devices are
blocked. The proposed cognitive use case assumes that either the SIM can be provisioned over the
air for properly authenticated users, such that the phone would function with priority access.
Alternatively, the system could determine that the user was a priority user based on the device ID
(as opposed to the priority access code in the SIM) and allow access that way as well; however, the
system computational effort to determine whether a call is being initiated by a priority user may
involve substantial computational requirements.
We recognize that assignment of priorities presents challenges in making the determination of
what communications are more important than others. Part of the ACCOLC decision criteria is the
understanding that implementing ACCOLC would deny access to the network for responders (or for
11
Note that concept of this use case would apply for other types of networks but reconfiguration would be
implemented in ways other than over-the-air provisioning of SIMs.
-24-
SDRF
SDRF-07-P-0019-V1.0.0
victims and observers who are providing critical information or notifying others). The ability to
prioritize communications as proposed in this use case does not guarantee that all critical or
important calls are made—there are physical limitations to the capacity of any system. However,
this use case provides the opportunity to utilize cognitive radio capabilities to implement the best
decisions that can be made with the available information.
In addition to the changes to dynamically modify user priority, it is also important to be able to
restore default conditions, such as when the user no longer requires priority access. Different
mechanisms for restoration may be implemented but could be similar to the same mechanisms used
to implement dynamic prioritization. Restoration could be executed based on a variety of
mechanisms, for example user request, incident command direction, and location if the user moves
out of the incident area. (see Section 5.4).
4.3.3.4 Regulatory Implications
The ability to define priorities and block access to the system for certain types of priorities is
part of the GSM specification. 12 Since this use case does not change the basic mechanism of
ACCOLC, the regulatory changes are limited to only those that may be necessary to allow
reprovisioning of SIMs over-the-air to change priorities.
4.3.3.5 Policy Implications
There are a number of policy changes implicit in this use case.
•
Policies for determining the circumstances under which an emergency responder would
be eligible to “upgrade” priority?
•
What information is required to authenticate the eligibility of the user to operate with
higher priority?
•
What procedure is followed if a user requests priority? Is there any human in that
decision loop?
•
Under what circumstances does the device’s priority revert to original level? Could
reversion be automatic based on responder location, or time frame? Could incident
command generate a broad directive (e.g., priority communications no longer required for
a particular sector/unit/area) that cognitive capabilities could then execute to restore
default priorities for all users?
•
While not necessarily applicable in the case of ACCOLC, a more general capability to
manage user priorities in real-time could also allow reducing default priorities of
responders if their role in the response is less critical. Under what circumstances would a
responder’s priority be reduced?
4.3.4. Use Case 4: Interface to Non-First Responders
Cognitive radios could allow non-first responders communications access to first responders in
specific situations in which the non-first responders are actively participating in the response, while
ensuring that mission critical public safety networks are not impacted.
12
Siemens Insight Consulting, “Communicating in a crisis – which technologies can be relied on?” 22 September
2006, available at http://www.continuitycentral.com/feature0394.htm
-25-
SDRF
SDRF-07-P-0019-V1.0.0
4.3.4.1 Summary of Scenario Situation
In terms of the aspects of public safety communications:
1. Physical: A fourth bomb was detonated aboard a bus near Tavistock Square. A group of
doctors were within walking distance of the explosion and the injured people. The doctors
arrived on the scene more quickly than first responders and therefore had more timely
information than the first responders. No specific prior planning had taken place to apply
these physicians as emergency response resources.
2. Network: The doctors did not have radios on any Public Safety net, but they did have cell
phones and landlines were available in the building. However, no dispatch organization
knew of doctors’ availability, and thus had no ability to initiate contact with them. Any
communication from the doctors had to come from 999 calls to dispatch, with information
then relayed to command centers.
3. Procedural: Since they were not part of a first responder organization, the doctors had no
authority to communicate on the first responder network.
4. Regulatory: Communication used established facilities.
5. Chronological: No prior planning had been done for the specific incident, but medical
personnel are aware of the legal implications of what they do in an emergency situation.
4.3.4.2 Description of the Use Case
In a mass casualty emergency there is a possibility that there are civilians that have the ability to
provide added benefit to the responses that are taking place by the public safety community. In
some cases these may be the only response available for an extended period of time. Thus it would
be advantageous to leverage this capability and to provide direction to the efforts being put forth.
This situation arose in this scenario when the bomb went off in a bus near Tavistock Square, as
there was a group of medical doctors meeting in a nearby building. Thus there were a number of
qualified medical personnel who were immediately available but were not tied into the incident
command communications. These trained medical personnel were not associated with an EMS
provider but were on the scene and able to provide qualified medical information regarding
casualties. While this was in many respects a fortunate coincidence, a well meaning “good
Samaritan” can also do more damage than good if they are unaware of the full situation. Thus the
challenge is to establish effective communications with non-first responders without negatively
impacting the incident command communications system and capabilities.
In today’s communication environment the average person carries as a minimum a basic cell
phone with the possibility of text messaging, photo and video capture and transmission. This
cognitive use case is an example of how communications capabilities could be adapted to most
effectively take advantage of situations in which non-first responder personnel are positioned to
play a role in the response.
We recognize that current concepts of operation and existing procedures do not generally
include linking first responder communications networks to non-first responder personnel
(regardless of their potential role in a response), and any change to such procedures cannot
compromise first responder communications.
Also note that the ability to appropriately link first responder communications with non-first
responder personnel can also apply where some first responders are only equipped with commercial
-26-
SDRF
SDRF-07-P-0019-V1.0.0
equipment. In Europe, for example, volunteers (such as a volunteer fire service) that are part of the
response may be equipped with commercial handsets rather than radios that access public safety
networks. In such cases, the capability to link them into a first responder network is a vitally
important capability.
A cognitive radio capability could allow them to link appropriately (upon proper authentication)
to coordinate their activities with public safety professionals as needed. The following provides a
view of how this scenario with cognitive radio could unfold:
1. Physical: No change from the physical situation described above.
2. Network: The initial doctor(s) on the scene would call 999 to report the explosion, and
identify themselves as doctors qualified to provide information on the medical status of
casualties. Dispatch, upon satisfaction that the caller could provide relevant information,
reconfigures the network (infrastructure, portable, or both) to allow the caller to
communicate directly with the appropriate emergency management medical coordinator,
incident command, and so on as dictated by policy. Once the doctor no longer needs to be
connected (i.e., first responder personnel arrive on the scene, the doctor begins performing
other functions, or all relevant information has been communicated), the network
reconfigurations are rescinded.
3. Procedural: Appropriate procedures would be in place to verify that the doctor was
qualified to provide the information. This could be accomplished by having medical
personnel pre-registered in some manner so that a dispatcher could authenticate the caller
(e.g., password, biometric, etc.) and ensure that the individual’s credentials already existed
in a registry. The doctor(s) would have communications capability as needed to the
appropriate organizations within the incident command structure. Procedures would also be
in place to establish voice communications channels that would not disrupt mission critical
incident command channels.
Given the presence of pre-registered individuals, an additional capability to be leveraged is
the ability to push information from dispatch or incident command (using some type of
notification system) out to pre-registered users requesting that they make appropriate contact
with the incident command staff for allocation and assignment. Location-aware cognitive
radios could also provide information to incident command to refine a notification
procedure. For example, upon indication of a problem at Tavistock Square, the medical
coordinator in dispatch would look at a map that indicates current deployment of medical
resources. The concentration of doctors would show up immediately.
From a detailed roster of doctors near the bus, the dispatcher would select an appropriate
number, and would send a short message to their mobiles asking if they can respond to a
bomb emergency in Tavistock Square. Doctors would be selected based on their
qualifications and specialties. The Cognitive element in the network would establish an adhoc response network. Each doctor who responds affirmatively would receive network
identification information (i.e., callsign).
4. Regulatory: Appropriate regulations would be in place to allow non-first responders to
communicate over the designated channels in emergency circumstances. Depending on the
communications capabilities used, this may or may not require regulatory changes.
-27-
SDRF
SDRF-07-P-0019-V1.0.0
5. Chronological: This use case would require that some pre-planning takes place that allows
doctors to establish the means by which they can be identified as such during an event (i.e.,
registration).
The above use case postulates that the communications is initiated by the doctors. A variation
of this use case considers a situation in which any interface with the first responder networks is
initiated by incident command. In this case we assume that a qualified medical person is not preregistered but has arrived on the scene of the incident. The individual calls 999; dispatch relays the
information through normal channels to incident command. The incident commander or appropriate
authority within the incident command structure determines that direct communications with the
individual is desirable, in which case the appropriate reconfigurations are executed. The scenario
situation would be described as follows:
1. Physical: No change from the physical situation described above.
2. Network: No change from the network situation described above.
3. Procedural: A doctor arrives on the scene of the explosion and calls 999. The doctor
explains the situation and his/her qualifications to provide more detailed assessment of the
medical condition of the victims to the call taker. The call taker obtains contact information
for the doctor and relays the information to incident command. Incident command
determines that direct contact is beneficial—the cognitive radio capabilities then establish
the appropriate communications network linkage between radios within the incident
command/first responder units and the doctor at the scene. Note that additional non-first
responders can be added to the network through the same process as needed and
appropriate.
4. Regulatory: Rules are required to describe exactly how such an ad-hoc network is to
perform and what channels they use.
5. Chronological: This use case would require that some pre-planning takes place that allows
doctors to establish the means by which they can be identified as such during an event (i.e.,
registration). Registration would also cover responsibilities and liabilities that are assumed
by the individual to perform such functions in an emergency. At time of notification of the
event, each Doctor would have the option of responding, or opting out. If they respond
affirmatively, they would become an on-site resource for incident command.
4.3.4.3 Functional Capabilities
The specific functional capabilities involved in this use case depend on the approach used to
implement it. If the implementation involves reconfiguring non-first responder radios to provide
them with a capability to communicate with the incident command/first responders, then the
functional capabilities include the ability to download a waveform and the ability for the non-first
responders’ radios to be reconfigured accordingly. If the implementation is based on infrastructure
linking in a non-responder radio, then there must be a means for the non-first responder radio to
upload information about the radio type.
A key element of this use, although not considered part of the cognitive capabilities, is ensuring
that any user who is linked to the first responders has a legitimate need for such communications,
and has a device that will not adversely impact first responder communications. Establishing that a
user has a legitimate need for such communications involves a number of issues:
-28-
SDRF
SDRF-07-P-0019-V1.0.0
•
Is the person someone appropriate for working with first responders? One approach
used currently for interaction between first responders and civilians is to utilize some
pre-registration process for translators, ministers, physicians, hazard experts, etc.
Background checks were performed on all individuals prior to use.
•
How is the identity of the user verified? Is the user who he/she claims to be?
•
How is information provided by a user verified? In the scenario, assuming a doctor
notified dispatch of the existence of casualties, how can that information be verified
to a level of confidence necessary to modify communications? (Note that dispatch
call-takers routinely evaluate the information provided in incoming calls.)
•
How is the potential role of a non-first responder verified?
In addition, the role of such responders would need to be incorporated into the incident
management (e.g., NIMS) as appropriate. Note that it is common for existing dispatch centers to
have a capability to patch a phone line to a radio channel. This cognitive use case extends that
concept to include establishing links between first responder communications channels and non-first
responder wireless devices/radios.
4.3.4.4 Regulatory Implications
Regulatory implications may also depend on implementation approach. Reconfiguring a nonfirst responder radio to be able to have some type of access to a public safety network would require
changes to the manner in which radios are currently type accepted. The alternative implementation,
in which the non-first responder communicates on their existing frequencies which is patched to a
public safety network frequency or channel is done with current technology and would not generally
require regulatory changes.
4.3.4.5 Policy Implications
Policy implications are dependent on the extent to which “non-first responders” are currently
incorporated into emergency response/incident management. In locations such as the United
Kingdom, where volunteer responders are incorporated into the response team, the only policy
change would potentially involve:
•
Guidance on the circumstances under which the cognitive capabilities would be exercised
to establish access for non-first responders (including policy on authentication, security,
and procedures);
•
Guidance on what communications are appropriate under such circumstances; and
•
Guidance on when and how non-first responders are disassociated with the network.
For agencies that typically do not utilize direct communications with non-first responders, such
as those in the U.S., any implementation of this cognitive capability would involve much broader
policies.
-29-
SDRF
SDRF-07-P-0019-V1.0.0
5. MULTI-USE CASE FUNCTIONAL CAPABILITIES
In addition to the functional capabilities that are identified for each specific use case in Section
4, there are functional capabilities that need to be incorporated into any cognitive use case (or
combination of use cases) to be operationally viable. In this section we identify those additional
functional capabilities, including: Role-Based Capabilities (Section (5.1), Command & Control
(Section 5.2) and Security (Section 5.3).
5.1. Roles
Many of the concepts of cognitive capabilities include a notion of the “role” of a responder.
Although not formally defined in this document, the general concept of role is the function being
performed by a responder: traffic control, perimeter security, fire suppression, logistics coordinator,
medical transport driver, etc. The role of an individual is typically assigned through a hierarchical
chain of command starting with an incident commander (or supervisor during normal operations).
An individual’s role may change during the course of an incident, and an individual may be
performing more than one role simultaneously. The individual’s communications capability needs
to adapt to his role assignments. The use case discussed in Section 4.3.4 postulates a situation in
which non-first responders are temporarily fulfilling roles.
Different roles require different communications capabilities, in terms of who needs to talk to
whom, the relative priority of communication, the quality of service required for the
communication, and so on. Communications capability should support the role of the radio user.
This concept exists in rudimentary form today—often departments have commander or supervisor
radios with additional functionality and/or channels/talk groups. However, current radio
configurations are defined statically such that they require reprogramming or hardware changes to
be reconfigured.
The definition of roles has two components: the role that the responder is performing in the
course of an incident; and the capabilities of the radio necessary to provide that responder with the
communications and connectivity to perform that role. The mapping of radio capabilities to
responder roles is generally a pre-defined assignment based on agency policies. A simple example
in place today is that some agencies provide supervisors with radios that are programmed differently
and/or have different capabilities than other officers. Role definitions could be implemented as a
policy that can be authenticated and downloaded once the role of the responder is adequately
established.
Cognitive radios can reconfigure based on who the user is and the role that the user is assigned
in the response. The ability of radios to be reconfigured based on roles addresses one of the public
safety concerns about reconfigurable radios—that radios that can be reconfigured to operate on
multiple bands and multiple channels could lead to chaos if “everyone can talk to everyone.”
Introduction of the concept of roles can allow department policies to govern communications
capabilities while permitting flexibility to allow the communications to evolve as incidents evolve
and responders assume various roles. Users performing multiple roles simultaneously could have
the union of capabilities associated with the roles.
Cognitive capabilities to support the concept of roles include the following:
•
Ability to define radio capabilities and connectivity as a function of roles.
-30-
SDRF
SDRF-07-P-0019-V1.0.0
•
Ability to interpret policies and procedures as a basis for cognitive decision-making.
•
Ability for a radio to support multiple roles.
•
Ability to reconfigure radio and network to support different roles in real-time.
•
Ability for an authorized user to define new roles and associated communications
capabilities in real time, and forward the information to the radios.
While the concept of roles provides a useful construct for reconfiguring radios, we also
recognize the challenge of linking the real world activity to cognitive radio capabilities and to the
command and control system (as discussed in Section 5.2). When responders are assigned a
responsibility within an incident command structure, it may be feasible to assign a role and
corresponding communications capabilities based on that role. However, roles may change in realtime, or roles may change based on events at the scene. Agency policy could dictate whether roles
could be changed by someone at a higher level in the command and control structure, by responders
themselves, or automatically as a function of data type leveraging cognitive capabilities. (Only a
small number of situations seem appropriate for the last case, for example, a “man down”
communication is assigned highest priority.) Establishing a workable policy that provides flexibility
to adjust priorities in real-time without burdening the command structure will require additional
research and thought.
5.2. Command and Control
One of the consistent themes in all of the cognitive use case discussions is that the flexibility
inherent in cognitive radios allows communications to be reconfigured to meet unexpected,
unanticipated aspects of the response. Each use case is rooted in situations for which the system was
not required to accommodate:
•
Destruction of infrastructure in the tunnel,
•
Overloading of the available communications capacity,
•
The priority of responders’ roles differing from static priority assignments in their
communications capabilities, and
•
Non-first responders having a potential role in the response.
Cognitive radios not only include the ability to reconfigure to adapt the communications to the
evolving situation but could also incorporate some decision making capability into the network and
subscribers as well.
One of the challenges unique to this particular scenario is that it evolved from four isolated
incidents which, due to their proximity in time and location, evolved into a major incident. One of
the incident management challenges was to configure the communications to support the rapidly
evolving command and control required to respond to the incident. Initial response procedures were
initiated for each incident and needed to be coordinated with a broader response that was executed
as the full scope of the attack became clear. Communications capabilities must be able to evolve as
an incident evolves—communications structures to support a single incident (such as one of the
individual bombings) must be rapidly expanded when coordination of responses to multiple events
is required.
-31-
SDRF
SDRF-07-P-0019-V1.0.0
To realize the benefits of cognitive radios in the public safety domain without inviting chaos, it
is therefore critical that:
•
Cognitive capabilities, and the decision making process, be under the control to the
extent desired of human(s) responsible for routine or incident response
communications (e.g., NIMS Comm Unit Leader, network manager). Note: “To the
extent desired” in the preceding statement allows a scalable array of cognitive
capabilities to be deployed with varying degrees of human interaction as needed. For
example, in the example of dynamic spectrum access, cognitive radios could be
allowed to utilize appropriate frequencies as needed, or could provide a
recommendation for human confirmation, or be directed to reconfigure to a new
frequency specified by a human.
•
Communications must support the appropriate command and control policies and
procedures governing an incident or routine response (e.g., NIMS).
Functional capabilities to support command and control and the functions of communications
management include the following:
•
Status information concerning subscriber radio configuration must be accessible by a
user, though an appropriate combination of audio and visual cues and visual displays.
•
Status information concerning radio and network configuration must be accessible by
an authenticated network manager.
•
System design incorporating cognitive capabilities should include options for human
intervention in the decision process.
•
Cognitive capabilities should be able to incorporate command and control policies
and procedures.
It is also worth noting the incident location in this scenario did not change, there are other
scenarios in which an incident (or collection of incidents) could spread or move—for example, a
riot, a kidnapping, or airborne release of a hazardous chemical. The public safety response needs to
be equally flexible and to take into account the incident location and the impact on its surroundings
– both ‘where it is now’ (current threat) and ‘where it has just come from’ (evidential retention)
while seeking to anticipate ‘where it might go next’ (public safety). In such scenarios it is vital that
any reconfiguration be under control of a central authorisation. While much of this can be
accomplished automatically, some degree of human interaction is required to (a) maintain situation
awareness of what is actually occurring and (b) to ensure post-Incident Management evidential
continuity and/or contingency planning. 13
Another aspect of control that cognitive radios could support involves the challenge of first
responders “self dispatching” or otherwise participating in the response without the knowledge of
incident command. Situations have occurred in which responders’ dedication and strong desire to
save lives results in participation in the response without properly coordinating with, or assignment
by, incident command. Responders, who have “self-dispatched” themselves to incidents need to be
identified, and either allocated to their current activities or withdrawn and reallocated to appropriate
duties. 14 Cognitive radio capabilities can support this process in two areas. First, cognitive
13
14
Neal Skelton and Mark Cartwright, private email correspondence.
Ibid..
-32-
SDRF
SDRF-07-P-0019-V1.0.0
capabilities of network attached radios could identify a peer radio that is not attached to the network
even if the non-attached radio is only using a peer-to-peer mode of transmission, or potentially some
other means of peer discovery. This capability could help identify the presence of responders who
are not properly authenticated into the network and may be unknown to incident command. Second,
once incident command is able to use the information to contact such a responder, they can be
assigned an appropriate role in the response and their radios reconfigured accordingly.
The ability of cognitive radio technology to provide tools for enhanced command and control
will require significant training for agencies to realize their benefits. We have postulated several
capabilities that give communications unit leaders, network managers, and so on significantly
greater control over the communications so that communications capabilities can adjust in real-time
to the evolving needs of an incident. Although there are many approaches which can be
implemented to realize these use cases, in general we have assumed that:
•
To the extent feasible cognitive capabilities can automate some aspects of the
communications configuration and reconfiguration while allowing human oversight
(i.e., not fully autonomous systems).
•
Cognitive capabilities should make as much of the reconfiguration seamless to the
end user.
•
Cognitive feature capabilities can be tailored to agency needs.
A key aspect of control, both at the network level and the subscriber level, is to ensure that
reconfiguration is done with the users’ knowledge and, as defined by policies and procedures, with
the users’ consent. While much of the benefit of cognitive capabilities involves making the behavior
of the radios and the network achieve functional goals without requiring significant user
intervention, at those times when radio behavior will change (temporarily being reconfigured,
changing priorities, connecting with non-first responders), the users need to be aware of the change
of capabilities. Policies and procedures will also need to be defined to clarify what control
individual users have over the cognitive capabilities. These policies may be defined in terms of the
role of the users as discussed in Section 5.1. An example of a policy that likely would be included
is the ability to override radio reconfiguration to ensure that life-critical communications are
transmitted (i.e., the man-down alarm always gets transmitted regardless).
Assuming the deployment of the capabilities outlined in this document, one of the challenges
will be developing sufficient numbers of trained personnel to manage the network, define
communications links to support incident command, and provide the appropriate human
intervention over the functions of the network. Pre-planning and coordination will continue to be
important, but the capability of cognitive radio technology to allow adjustment of communications
capabilities in near real time adds complexity to the communication management. Also, capabilities
that minimize the demands on the end user are likely to add complexity to the overall network
management. In the NIMS structure in the United States, this function is the responsibility of the
Comm Unit Leader. For the Gold Coordinating Group, there will be someone responsible for
managing network configuration and reconfiguration, authorizing access to additional spectrum, and
other similar functions. It will be necessary to provide such individuals with education, training,
tools, and experience (e.g., through exercises, participating in support functions during actual
incidents) to prepare them to execute this responsibility.
-33-
SDRF
SDRF-07-P-0019-V1.0.0
5.3. Interoperability
The After Action reports for this scenario did not specifically identify communications
interoperability as a significant impediment to the response to the bombings. The initial phases
included resources from 3 separate police forces – Metropolitan Police, British Transport Police and
City of London Police forces. All are separate and distinctly individual police forces that have their
own command/reporting chains. In addition there were responders from the London Ambulance
Service involved as well. Communications between responders of these organizations typically only
occurs at the highest level. However, by linking cognitive radio technology to the concept of roles
described in Section 5.1, there may be opportunities to enhance interoperability among responders
based on the communication requirements of the incident. For example, if based on roles it was
useful for responders from one agency to communicate with responders from another agency,
cognitive radio capabilities could either reconfigure the radios or identify a common channel to
support interoperability.
Cognitive radio capabilities could also simplify the planning associated with a major event. For
example, when additional law enforcement personnel are brought into an area to assist with an event
(e.g. 300 officers from West Midlands Police drafted in to help out at the UK’s Millennium Stadium
in Cardiff on Football Cup Final day) it would be useful if their “home” (Wmids) radios could be
reconfigured temporarily on the Gwent network.15
5.4. Restoring Default Configurations
Each of the use cases described in Section 4 involve dynamically reconfiguring subscriber
and/or network capabilities based on the evolving requirements of a dynamically changing incident.
At some point as the incident winds down, the capabilities must be reconfigured to return to their
normal default capabilities. To some extent the concepts that have already been identified, such as
roles, can be applied to the ramp down of the incident as well as to the ramp up—when a responder
reverts to their normal (daily operations) role, the associated communications capability reverts as
well. For example, priorities in a dynamic prioritization scheme would be restored to their normal
levels. In the case of network extension, once the responder is back within the normal coverage
footprint of the system (e.g., out of the tunnel) the radio would be reconfigured to its initial
configuration.
To achieve this ramp down gracefully, there are additional functional capabilities that would be
useful in addition to those identified in the use cases in Section 4:
•
15
A capability to recognize when the circumstances that caused initial reconfiguration
are no longer relevant. In the case of radios being out of network coverage, the
circumstance would be when the radio is back within coverage. In the case of
dynamic prioritization, the circumstance would be when the role of the responder
changes. In some cases, such as the dynamic prioritization based on responder role,
there is a discrete point in which one can explicitly identify that the circumstance has
changed—the responder is being discharged from the incident response assignment.
In other cases a discrete point will be more difficult to identify—for example, when
the demand for communications resources has declined to the point where additional
Ibid.
-34-
SDRF
SDRF-07-P-0019-V1.0.0
spectrum is no longer required. In these cases, specific cognitive functions may be
required to identify when the default configurations should be reinstated.
•
Records of the configuration changes, both for audit and post-incident analysis
purposes, as well as to specific “undo” configuration changes if necessary.
•
A cognitive capability to determine whether preferable to simply restore defaults or to
sequentially undo each change.
5.5. Security
Cognitive radio technology requires consideration of potential risks to public safety
communications to avoid introducing new vulnerabilities. By expanding communications options
available to first responders, cognitive radio technology also has the potential to open new points of
vulnerability in public safety communications systems. For instance, attempts to exploit new
cognitive features by individuals or organizations in order to compromise the availability,
confidentiality, or integrity of first responder communications is almost certain. Fortunately,
technical design features and operational controls should be able to mitigate potential risks to the
level that they do not represent a significant threat to the public safety mission. This section briefly
reviews three potential controls (user and role authentication, device and network authentication,
and reconfiguration control) and recognizes associated technical, operational, and regulatory
changes that may be required to implement them.
5.5.1. User and Role Authentication
Many advanced public safety communications systems support user (operator) authentication
today to prevent unauthorized individuals from accessing public safety networks. However, user
authentication in these systems is typically limited to public safety personnel operating on their own
closed network. Thus anybody who can access a radio can access and possibly disrupt
communications on these networks. In contrast to today’s situation, possibilities exist for non-first
responders to access the network in one of the cognitive use cases. In another use case, first
responders are granted higher priority on a non-public safety network, specifically one supporting
commercial cellular telephony. In both instances, there is a need to authenticate users seeking
privileges not common in day-to-day use of the system. Without a more robust authentication
method , in each instance the potential exists for significant misuse or abuse of the cognitive
functionality. This can result in congestion or the prevention of communications essential to the lifecritical emergency situation.
Authenticating external users to an otherwise closed network need not be a significant
challenge, as there are a variety of solutions available from which to choose depending upon the
circumstances involved. For example, if the user who needs to join the network has a compatible
radio then it might be a simple administrative action over the air to identify the user to the network
by a known individual who is already a part of the network. This is known as authentication by
proxy. This method can work in a variety of other circumstances. For instance, there may be an
individual who is on the scene who needs to get priority on the public cell phone network. A known
individual can obtain the ESN of the user and pass it into the public safety system and have the
user’s cell phone granted a temporary priority until it is no longer needed. Another example that
external users have to add or upgrade authentication software or other supporting software
-35-
SDRF
SDRF-07-P-0019-V1.0.0
components on their devices to interoperate with the target network. A cognitive radio could
possess the capability to ascertain what components are needed and, with the users authorization,
the cognitive radio could automatically proceed to download and implement the necessary software.
In other instances, users likely will need to complete some form of registration process prior to the
actual use of the system. The managers of the network will also need to build directory services and
associated databases to support authentication transactions for the users of the system. In some
cases, regulatory changes may be required if current public safety administrative regulations require
security controls that are incompatible with the proposed technology, or if use of a network requires
personnel clearances not held by the users that would have access to the network. Other regulatory
issues might arise if the terms of allocations and radio licenses limit operation to certain classes of
users.
One way to simplify some of the complexity resulting from cognitive capabilities is to limit
authentication to the user’s role rather than the user’s identity. For example, in a major emergency
such as the London bombing, it likely is more important that first responders know that a network
access request is coming from a medical doctor (a role) than it is to know that the request is from
Dr. John Smith of Central Municipal Hospital (a user). A role-based system may need to handle
from a limited number of roles, while a user-based system may need to manage thousands of users.
Individual users still need to authenticate themselves as being legitimate users, but this
authentication can be managed with simple schema currently used (e.g. passwords or physical
tokens) in such cases.
To support a role-based system, a user in a particular role could obtain a public key certificate
from a nationally trusted certification authority that he or she legitimately serves in that role. 16 The
technical characteristics of public key certificates would enable the user to authenticate in any
jurisdiction that recognized the national authority, thereby obviating the need for each jurisdiction to
maintain or connect to databases for all potential users in that role. Certificates could also be
accompanied by relevant capabilities of the individual, such as that he or she is a trauma specialist
or has expertise in the handling of explosive or radioactive material. An important consideration
with the use of certificates is that a method is required to securely store the credentials. Otherwise
they can be exploited by anybody. Maintaining the credentials in the radio (or even in an external
token) implies that they can be protected from physical attack and that the actual user has a means
of unlocking the certificate and authenticating himself to the radio, such as a user ID and password.
While technology to support this system is widely available, the cognitive radio abilities described
would require the set-up of a supporting public key infrastructure and related policies and
procedures for this particular application.
In some cases, ad hoc roles may be established to support a particular incident. For example,
emergency responders may want to identify users that carry camera phones to get images from
locations that might be inaccessible as a result of a disaster. In this case, the cell phone users would
not be members of a pre-registered group such as medical personnel, but still should be
distinguished from users that are not supporting incident response. Users that can serve in a
supporting role could be authenticated based in part on their location or, as described earlier, by
16
Another mechanism to achieve role-based authentication is to employ call group keys that are used to encrypt
message content. Members of a call group could be associated with a role. When keys are used to encrypt
communication, they serve as a de facto form of authentication because those without the relevant key will be
unable to interpret encrypted communications. PKI provides the tools for key management to ensure that all users
who require keys have them..
-36-
SDRF
SDRF-07-P-0019-V1.0.0
proxy Cognitive technology could be used to identify these users and determine how to best
differentiate them on network so that incident command can easily access them
5.5.2. Device and Network Authentication
With the exception of including ad hoc users into a situation, and as long as the user
authenticates himself to the device, it is not the user that needs to be authenticated to the network;
rather the device that the user carries. For example, one use case for cognitive radio involves
extending the network to radios that cannot reach the regular infrastructure for whatever reason. In
this case, the cognitive radio, in effect, serves as a repeater or interface bridge for the radios for
which the infrastructure is unavailable. The new functionality introduces two important security
issues: first, how do we know that the first responder’s radio is connecting to the cognitive repeater
and not a rogue radio, perhaps one setup by those who caused the emergency and disruption in
communications; second, how is the system preventing rogue radios from accessing the network via
the cognitive repeater?
For situations that do not involve legacy radios, the solution is to support mutual authentication
between the device and the network to which it connects. In the network extension case, the
cognitive radio that extends the network should support an authentication pass-through that would
enable authentication services to be performed just as it would under normal circumstances. Of
course the cognitive radio which is providing the pass through would first have to authenticate itself
to the network. Techniques such as Extensible Authentication Protocol (EAP) support this
capability today and are implemented on Wi-Fi equipment and other communications systems.
Moreover, they can be implemented in a manner than is transparent to users and does not require
operational or regulatory changes. Nevertheless, new cognitive radios that support network
extension should be accompanied by device and network authentication modules to protect systems
against attack.
5.5.3. Reconfiguration Controls
One of the use cases for cognitive radios is to switch protocols to balance load or allow
additional access. In this case, the radio automatically reconfigures itself to provide new benefits.
The risk is the radio will not perform the intended function after reconfiguration, perhaps because an
adversary has modified radio software modules to cause a system failure or connect to an
unauthorized network.
As with any reconfigurable radio, cognitive radios will need to implement security controls to
provide assurance that reconfiguration transactions perform as expected with very high integrity and
sufficient robustness. Code signing, RF transmission policy enforcement, and memory isolation
using secure operating system kernels are all currently technology mechanisms that can prevent
unauthorized behavior. The SDR Forum’s Securing Reconfigurable Communications Technology
document (forthcoming) explains these potential controls in greater detail.
5.6. Transition and Interface to Legacy System
Realization of any of the use cases described above will not happen “overnight” even at the
point where the technology is proved, supporting regulatory changes have been instituted, and
-37-
SDRF
SDRF-07-P-0019-V1.0.0
policies and procedures have evolved accordingly. Thus in each use case functional capabilities,
regulatory considerations, and policies and procedures will need to accommodate legacy systems as
well.
Specific functional capabilities required to realize this capability include the following:
•
The network needs to be able to distinguish legacy equipment and the capabilities of
each device—note that since legacy equipment may not be programmed or
configurable to transmit specific capabilities information to the network, it may be
necessary for the network to recognize the waveform characteristics and derive/infer
the device capabilities.
•
The network needs to be able to reconfigure transmit/receive capabilities to interface
with legacy equipment and bridge as appropriate to establish necessary
communications with other subscribers on the network.
•
The network may also need to establish/re-establish communications links to avoid
interference from legacy equipment.
•
An appropriate subset of the above functions should also be implemented at the
subscriber level.
5.7. Interaction of Use Cases
While these use cases have been described and analyzed independently, there are interactions
among the use cases and the cognitive capabilities that support the use cases. For example, dynamic
spectrum access and dynamic prioritization could interact such that responders with the highest (or
lowest) prioritization could utilize dynamically allocated spectrum to maximize access for the
highest priority users. Dynamic spectrum access techniques may be useful in realizing the network
extension use case. Network extension capabilities could be used to link to non-first responders. We
also recognize that the issues of command and control described in Section 5.2 become even more
challenging as capabilities are implemented to realize multiple use cases.
It is beyond the scope of this report to address all of the issues that could arise in the interaction
of use cases, so we simply note that there are additional interactions to consider in addressing
multiple use cases, in terms of functional capabilities, regulatory considerations, and policies and
procedures.
5.8. Standards
Some of the use cases involve communication of control information among subscribers (e.g.,
discovery of peer radios) or between the network and subscribers. Examples of the latter include
information concerning roles and priorities. There needs to be standard protocols for the
transmission of such information to ensure that radios are interoperable.17 For example, in the
network extension case (Use Case #1), a result in which Vendor A’s radios cannot create a link with
Vendor B’s radios would set back much of the ongoing effort to develop standards such as P25 and
17
Recent legislation passed by the U.S. Congress requires that federal grant money can only be spent for
communications equipment developed pursuant to voluntary consensus standards where such standards exists.
-38-
SDRF
SDRF-07-P-0019-V1.0.0
TETRA that promote interoperability. The specific protocols requiring standard definitions will
depend in many cases on the specific approach used to implement functional capabilities identified
in the use case, so it is important to recognize that standards will be needed to ensure
interoperability of cognitive radios.
-39-
SDRF
SDRF-07-P-0019-V1.0.0
6. SUMMARY
The analysis in the preceding sections outlines significant capabilities that can enhance the
ability of public safety agencies to communicate, particularly under the challenging conditions of a
major event or incident. We have identified the potential use of cognitive capabilities to:
1. Extend existing network coverage when individual radios move outside the coverage
footprint of the communications system;
2. Dynamically allocate spectrum to provide greater capacity for overloaded networks;
3. Dynamically prioritize communications to better manage load; and
4. Dynamically reconfigure networks to incorporate non-first responders who have
information of value to incident management and/or responders.
In order to realize these use cases, there are a number of functional capabilities that need to be
developed (and tested and proven), including the following:
•
At the network level:
o Identify loading that meets criteria for initiating dynamic spectrum access.
o Identify spectrum resources that can be used to offload calls or expand capacity,
either by executing established agreements or by searching for underutilized
spectrum (“white space”) that could be dynamically accessed.
o Manage the hidden node problem.
o Deconflict the situation of multiple users attempting to access the same available
spectrum.
o Assign and download priorities to subscriber equipment based on user roles.
o Download reconfiguration information to subscribers.
o Recognize legacy equipment and manage communications accordingly.
o Restore normal and default configurations.
•
At the subscriber level:
o Determine when they are disconnected from the network.
o Find and identify peer radios.
o Create a network extension route to the infrastructure from each “disconnected”
device (also requires determination of non-interfering frequencies for each
“hop”).
o Adjust network as devices enter and exit the area.
o Preserve the level of security of the baseline network in the network extensions.
o Reconfigure to incorporate the added transmit/receive frequencies.
o Reconfigure to function as repeater.
o Change priority levels.
-40-
SDRF
SDRF-07-P-0019-V1.0.0
o Interaction with user and device authentication schemes.
o Ensure reconfiguration does not compromise the security or integrity of the radio
or its operations.
We recognize that some these capabilities are deployed today in some form, while others may
require significant development. We also recognize that there are substantial regulatory and policy
and procedural considerations and changes that will be required in conjunction with the technical
developments to realize the impacts described in Section 4. However, we also realize that
development of cognitive capabilities can dramatically increase the ability of incident commanders.
First responders will be assured that critical information will flow as needed despite changes in
coverage, connectivity, and loading on communications systems.
-41-
Public Safety Special Interest Group
Use Cases for Cognitive Applications in Public Safety Communications Systems: Volume 1
SDRF-07-P-0019-V0.5
1
APPENDIX A: LONDON BOMBING SCENARIO
2
3
4
This Appendix provides the timeline events during the London bombing incident of July 7, 2005. In the Scenario there are a number
of places in the timeline (use cases) where the PS-SIG felt that SDR capabilities could provide an improved capability to what is in place.
In the timeline these events are shown with a green shaded cell in the table. Section 4.3 provides an expansion on these use cases.
5
6
7
The ⌧ in the Communication Activity is an indication that there is a form of electronic communication taking place. This
could be in the form of RF or data via IP.
8
Table 2. London Bombing July 7, 2005 Timeline of Events
Time/
Event
8:50 AM
First bomb
detonated
at
Liverpool
St.
8:51 AM
Second
bomb
detonated
at Kings
Cross.
8:52 AM
PS Response
On-scene Problems /
Situation Awareness
Communication
Activity
Multiple, often conflicting, reports
were being made, some to London
Underground’s Network Control
Centre, some to the emergency
services, and some to the media.
There were reports of loud bangs.
There was a loss of power on sections
of the Underground. 999 calls were
made from nearby locations reporting
smoke issuing from tunnels and from
a grid in a street close to Edgware
Road. It was not clear what had
happened, or indeed where.
Serious and non-serious injuries
result.
In the minutes following the
explosions on the Tube trains,
passengers were plunged into total
darkness.
999 calls were made from
nearby locations.
{Units dispatched.}
A-1
Potential Cognitive Capability
Line control centers identify lack of
communications, attempt alternative
routing, generate alert. (Note that
cognitive capabilities in the train radio
system could be implemented to
provide this capability, but since it
Public Safety Special Interest Group
Use Cases for Cognitive Applications in Public Safety Communications Systems: Volume 1
SDRF-07-P-0019-V0.5
Time/
Event
PS Response
On-scene Problems /
Situation Awareness
Communication
Activity
They did not know whether anyone
knew they were there, or if help was
on its way.
The internal carriage lights went out,
internal communications between the
driver and passengers of each train
were debilitated, and
Drivers were unable to communicate
with their line control centres.
Passengers on the three bombed
trains were unable to communicate
with the drivers of the trains to alert
them to the explosion
8:53 AM
8:54 AM
Third
bomb
detonated
at Edgware
Rd
8:55 AM
8:56 AM
8:57 AM
8:58 AM
does not involve public radio systems
it is not included in the use cases in
Section 4.
The first British Transport Police
officer arrives at the scene and
reported ‘building shock’ and
smoke issuing from the tunnel, but
no evidence of structural damage.
King’s Cross The Metropolitan
Police Service was first alerted to
an incident at King’s Cross on the
basis of CCTV footage of the
station.
⌧
Aldgate The British Transport
Police had identified the site of the
incident in the tunnel between
Aldgate and Liverpool Street, but
had not discovered any injured
passengers at that point. Power to
⌧
⌧
A-2
Potential Cognitive Capability
Public Safety Special Interest Group
Use Cases for Cognitive Applications in Public Safety Communications Systems: Volume 1
SDRF-07-P-0019-V0.5
Time/
Event
PS Response
On-scene Problems /
Situation Awareness
Communication
Activity
the track was cut off.
Edgware The British Transport
Police receive a call to Edgware
Road, reporting a person under a
train and a train collision with the
tunnel wall.
8:59 AM
9:00 AM
9:01 AM
9:02 AM
9:03 AM
9:04 AM
Network Control Centre
put in an emergency
services call to three sites
– Aldgate, King’s Cross
and Edgware Road.
Aldgate The first fire engines
arrive. Further Fire Brigade units
were mobilized to a reported
explosion at Aldgate.
Edgware Road The London Fire
Brigade mobilized five units,
including a Fire Rescue Unit and a
Fire Investigation Unit, to Praed
Street.
⌧
Further appliances were
mobilized, responding to reports
of smoke in a tunnel. Two fire
engines and a senior officer were
sent to Aldgate, and an additional
fire engine was sent to Liverpool
Street.
King’s Cross The London Fire
Brigade received its first 999 call,
reporting smoke issuing from a
tunnel at King’s Cross.
Aldgate The first ambulance
arrives at Liverpool Street,
Edgware Road: The first London
Fire Brigade units arrived at Praed
Street.
⌧
Praed Street turned out not to have
been the site of any incident
A-3
The Metropolitan Police
were called by the
London Fire Brigade and
Potential Cognitive Capability
Public Safety Special Interest Group
Use Cases for Cognitive Applications in Public Safety Communications Systems: Volume 1
SDRF-07-P-0019-V0.5
Time/
Event
PS Response
On-scene Problems /
Situation Awareness
King’s Cross At 9.04 am, a ‘split
attendance’ was mobilized, with
three fire engines sent to Euston
Square and one to King’s Cross.
9:05 AM
9:06 AM
9:07 AM
9:08 AM
9:09 AM
9:10 AM
9:11 AM
Emergency planning manager
arrives at Liverpool Street.
The London Ambulance Service
Emergency Planning Manager
advised Central Ambulance
Control to place hospitals on
major incident standby, identify
safe rendezvous points in case of a
Chemical, Biological, Radiation or
Nuclear (CBRN) risk, and
mobilize equipment vehicles.
Edgware Road Fire Control
received a call alerting them to the
location of the incident on the
Hammersmith and City Line at
Edgware Road station.
Fire engines arrived at Euston
Square (which turned out not to be
one of the sites where passengers
were emerging from tunnels).
The British Transport Police at the
scene reported that there had been
a train accident, and declared a
major incident.
Communication
Activity
were on the scene at 9.12
am.
There are 25 walking wounded, some
of whom were badly injured.
The City of London Police
recognized that there had been an
explosion caused by a bomb, and
declared a major incident.
Additional Fire engines arrived at
Euston Square (which turned out
not to be one of the sites where
A-4
⌧
Potential Cognitive Capability
Public Safety Special Interest Group
Use Cases for Cognitive Applications in Public Safety Communications Systems: Volume 1
SDRF-07-P-0019-V0.5
Time/
Event
9:12 AM
9:13 AM
9:14 AM
9:15 AM
PS Response
passengers were emerging from
tunnels).
Edgware Road The first
ambulance arrives at Edgware
Road.
Edgware Road Metropolitan
Police on the scene.
King’s Cross The first fire engine
arrived at King’s Cross station.
Fire Control – four vehicles were
mobilized to Edgware Road. Only
one of these was a redeployed
vehicle from Praed Street.
Aldgate An ambulance crew
reported that the incident had been
an explosion, and that there were
five fatalities.
Edgware Road Ambulance corps
confirmed that there had been an
explosion and requested ‘as many
ambulances as you can muster’.
The crew reported back to the
control room that there had been
an explosion with up to 1,000
casualties
King’s Cross A London
Ambulance Service Fast Response
Unit arrives at King’s Cross
The London Fire Brigade declared
a major incident.
Kings Cross: A major incident
was declared at King’s Cross by
On-scene Problems /
Situation Awareness
Communication
Activity
Communication between the control
rooms of the emergency services in
the event of a major incident takes
places through a ‘first alert’ system.
This is done through a ‘first alert’
call, which is in effect a conference
call involving the emergency and
transport services.
The ‘first alert’ system was activated
at 9.12 am.
⌧
⌧
⌧
The London Ambulance Service was
initially called to seven separate sites,
and ambulances were being deployed
to ‘various places that ended up not
A-5
⌧
Potential Cognitive Capability
Public Safety Special Interest Group
Use Cases for Cognitive Applications in Public Safety Communications Systems: Volume 1
SDRF-07-P-0019-V0.5
Time/
Event
PS Response
the Metropolitan Police Service.
9:16 AM
9:17 AM
9:18 AM
On-scene Problems /
Situation Awareness
being the main incident sites’. For
some time, it was thought that there
may have been up to five separate
incidents on the Tube, and the
emergency services were being
deployed accordingly to five separate
Tube station.
A ‘leaky feeder’ cable that enables
the British Transport Police’s radios
to function was damaged by the blast.
Emergency and transport services
personnel were therefore unable to
communicate with their colleagues at
ground level without making the 15minute journey back down the tunnel
to the platform.
The decision was taken at 9.15 am to
declare a network emergency and
evacuate the entire Tube network.
⌧
On arrival at the affected trains,
emergency services personnel
sought to establish what had
happened, and needed
immediately to communicate this
information to control centres. The
British Transport Police is the only
emergency service equipped with
radios that can function
underground. All the other
emergency services had to rely on
individuals running back and forth
from the train to the platform and
from the platform to ground level,
or use British Transport Police
radios.
Edgware Road The first fire
engine arrived at Edgware Road.
Communication
Activity
Russell Square. The British
Transport Police reported that there
A-6
⌧
Potential Cognitive Capability
Public Safety Special Interest Group
Use Cases for Cognitive Applications in Public Safety Communications Systems: Volume 1
SDRF-07-P-0019-V0.5
Time/
Event
PS Response
9:19 AM
The British Transport Police
formally requested assistance from
the Metropolitan Police Service
(which is the lead police service in
the event of a major or
catastrophic incident, even if it
takes place within the jurisdiction
of the City of London Police or
British Transport Police).
King’s Cross First ambulance on
scene.
King’s Cross Further fire engines
were requested to King’s Cross.
9:20 AM
The Metropolitan Police was in
fact already aware of the incident,
and the first officer arrived at the
scene
King’s Cross A major incident
was declared at King’s Cross by
the London Ambulance Service.
9:21 AM
9:22 AM
9:23 AM
9:24 AM
On-scene Problems /
Situation Awareness
were at least 200 casualties.
It is unclear precisely when the
London Fire Brigade became aware
that there had been an explosion at
King’s Cross. However, we do know
that the ability of the London Fire
Brigade to establish what had
happened at King’s Cross was
hampered by the fact that hand-held
radios did not work effectively
between the platform and a control
position at the top of the escalator,
nor between the top of the escalator
and outside the station. The Fire
Brigade therefore had to use runners
– individuals running up and down
escalators – to communicate from
below ground to the surface.
Communication
Activity
⌧
Potential Cognitive Capability
Network extension for coverage &
reachback (see Section 4.3.1)
⌧
Russell Square The London
Ambulance Service dispatched a
Fast Response Unit.
LAS dispatched Fast
Response Unit.
A-7
As part of the analysis performed for
this document, it is noted that similar
information arrived at the BTS and the
LAS command centers at significantly
different times. After considering
potential cognitive capabilities to
improve the timeliness of information
flow, we concluded that the issues
primarily command and control issues
(procedural environment) rather than
Public Safety Special Interest Group
Use Cases for Cognitive Applications in Public Safety Communications Systems: Volume 1
SDRF-07-P-0019-V0.5
Time/
Event
PS Response
On-scene Problems /
Situation Awareness
Communication
Activity
Potential Cognitive Capability
communications issues. Cognitive
capabilities could support real-time
changes in the network environment to
support changes in command structure.
9:25 AM
9:26 AM
9:27 AM
9:28 AM
9:29 AM
9:30 AM
9:31 AM
9:32 AM
9:33 AM
9:34 AM
9:35 AM
9:36 AM
9:37 AM
9:38 AM
The first conference call
on the ‘first alert’ system
took place.
Russell Square. A Fast Response
Unit arrived at the scene.
Edgware Road The Metropolitan
Police Service declared a major
incident
⌧
Edgware Road The Fire Brigade
declared a major incident at
Edgware Road station.
⌧
King’s Cross Further fire engines
were requested to King’s Cross.
Edgware Road The Fire Rescue
Unit that had been sent to Praed
Street re-deployed to Edgware
Road
Russell Square A major incident
declared at Russell Square by the
London Ambulance Service (note:
this is 45 min after the explosion).
⌧
⌧
Russell Square The Ambulance
Service Professional Standards
Officer at the scene was reporting 615 fatalities and 50+ casualties
fatalities - and stated that there was
only one ambulance at the scene,
along with the Fast Response Unit.
Note: This was a full 20 minutes after
the British Transport Police received
A-8
⌧
Public Safety Special Interest Group
Use Cases for Cognitive Applications in Public Safety Communications Systems: Volume 1
SDRF-07-P-0019-V0.5
Time/
Event
PS Response
reports of loss of life and limbs.
Kings Cross. The ambulance crew
reported that there was still no officer
at the scene, but that there were 400
casualties and 15 ambulances were
needed.
Russell Square. The Metropolitan
Police Service requested the London
Ambulance Service to ‘send every
unit that you have got’.
9:39 AM
9:40 AM
9:41 AM
9:42 AM
9:43 AM
9:44 AM
9:45 AM
9:46 AM
9:47 AM
Fourth
bomb
detonated
9:48 AM
On-scene Problems /
Situation Awareness
Communication
Activity
Potential Cognitive Capability
⌧
⌧
Kings Cross. The first LAS manager
was sent to the scene, almost an hour
after the explosion.
⌧
Tavistock Square. It was
immediately apparent what had
happened, and the first 999 call
was made, within a minute of the
explosion. Twelve further 999
calls were made, all before 9.56
am. A number of medics were on
the site before that time: the bus
was located outside the
headquarters of the British
Medical Association and doctors
and other trained first-aid
personnel came out of the building
to care for the injured.
Russell Square. One ambulance
was dispatched from University
College Hospital.
⌧
9:49 AM
A-9
Interface to non-first responders (see
Section 4.3.4).
Public Safety Special Interest Group
Use Cases for Cognitive Applications in Public Safety Communications Systems: Volume 1
SDRF-07-P-0019-V0.5
Time/
Event
PS Response
9:50 AM
Fire engines dispatched to
Tavistock Square.
9:51 AM
9:52 AM
9:53 AM
9:54 AM
9:55 AM
9:57 AM
9:58 AM
9:59 AM
10:00 AM
10:01 AM
10:02 AM
10:03 AM
10:04 AM
10:05 AM
10:06 AM
10:07 AM
10:08 AM
10:09 AM
10:10 AM
10:11 AM
10:12 AM
10:13 AM
On-scene Problems /
Situation Awareness
Communication
Activity
⌧
Tavistock Square. The
Metropolitan Police Service
happened already to have an
officer at the scene.
The first ambulance arrived on the
scene at Tavistock Square, having
come across the explosion (as
opposed to having been
specifically dispatched there).
⌧
Russell Square. A request was
made for five ambulances and a
bus.
Kings Cross. duty officer reported
that there were still more than 50
casualties in the train, and requested a
further ten ambulances and an
A-10
⌧
Potential Cognitive Capability
Public Safety Special Interest Group
Use Cases for Cognitive Applications in Public Safety Communications Systems: Volume 1
SDRF-07-P-0019-V0.5
Time/
Event
PS Response
On-scene Problems /
Situation Awareness
Communication
Activity
Potential Cognitive Capability
⌧
Interface to non-first responders (see
Section 4.3.4).
equipment vehicle
Russell Square. The manager at the
scene reported that there were 40-50
walking wounded and 100 stretcher
cases still in the tunnel. There was
still only one ambulance on the scene
at that point.
10:14 AM
10:15 AM
10:22 AM
10:27 AM
10:30 AM
London Bus Control centre called
to request that walking wounded
be sent to Bart’s
Russell Square. An equipment
vehicle was requested.
Russell Square. The manager at
the scene requested an estimated
time of arrival of the ambulances
that had been requested. There
was no reply from Central
Ambulance Control.
Kings Cross. Four busloads of
casualties were taken (by bus drivers
who had taken the impressive
individual initiative of offering their
services) to The Royal London
Hospital. They were directed to the
Royal London Hospital, despite a call
to the control centre seven minutes
earlier requesting that walking
wounded be sent to Bart’s instead.
Kings Cross: No further information
was recorded about the Ambulance
Service’s response at King’s Cross,
other than the time at which the scene
was cleared of casualties – 2 hours
and 26 minutes after the explosion.
A system exists to restrict mobile
phone network access to the
emergency services within a specified
area. This system, called the Access
Overload Control (ACCOLC) is seen
very much as a last resort. It is
expensive to implement and can
cause public distress or panic. The
decision to activate ACCOLC can
A-11
Kings Cross: The
London Ambulance
Service manager at the
scene reported that there
were still 50 people
trapped in the train.
Several possible CR-based approaches
to improve response:
o Temporarily reconfigure
responder radio priorities (see
Section 4.3.2).
o Dynamically access additional
spectrum.(see Section Error!
Reference source not found.).
Public Safety Special Interest Group
Use Cases for Cognitive Applications in Public Safety Communications Systems: Volume 1
SDRF-07-P-0019-V0.5
Time/
Event
PS Response
On-scene Problems /
Situation Awareness
therefore be taken only at the highest
level of command: the Gold
Coordinating Group.
The first meeting of the Gold
Coordinating Group, considered
whether to close down mobile phone
networks to the public at any of the
sites where the emergency rescue
effort was being mounted. The
London Ambulance Service told us
that problems with mobile phones
and radios led them to as the Gold
Coordinating Group to activate
ACCOLC in the area around Aldgate
station, and that their request had
been refused by the Gold
Coordinating Group. It was decided
that ACCOLC should not be
activated, because of the risk of
public panic and also because it was
not clear that the right personnel
would be carrying ACCOLC-enabled
telephones. 18 If they were not
carrying this equipment, ACCOLC
could have made matters worse. As it
was, at least some mobile telephone
calls were getting through some of
the time. Had ACCOLC been
activated, key personnel who were
not carrying specially-enabled
telephones would not have been able
to make or receive any calls. This is
clearly a major flaw in the system:
there is no point in having the
technology to enable key people to
18
Transcript of Committee meeting, 3 November 2005, Volume 2, pages 24-25
A-12
Communication
Activity
Potential Cognitive Capability
Public Safety Special Interest Group
Use Cases for Cognitive Applications in Public Safety Communications Systems: Volume 1
SDRF-07-P-0019-V0.5
Time/
Event
PS Response
On-scene Problems /
Situation Awareness
Communication
Activity
Potential Cognitive Capability
communicate with each other if the
relevant authorities do not make sure
that the right people are in possession
of that technology.
10:45 AM
11:00 AM
11:15 AM
11:30 AM
11:31 AM
11:45 AM
12:00 AM
Tavistock Square. The tactical, or
‘Silver’, officer at Tavistock Square
reported that they had enough
vehicles. It turned out that this was
the result of ambulances destined for
Russell Square being directed to the
same muster point as those
dispatched to Tavistock Square.
⌧
ACCOLC had been activated, by the
City of London Police, on the O2
network in a 1km area around
Aldgate Station. This was a response
to the fact that the City of London
Police were experiencing serious
communications difficulties in the
area, and this was hampering their
response. Despite the Gold
Coordinating Group decision, the
City of London Police made a request
to O2 to shut down the O2 network to
the public in a 1km area around
Aldgate station. O2 carried out the
appropriate validation procedures, but
these procedures, set by the Cabinet
Office, do not include verifying the
request with the Gold Coordinating
Group. The O2 network was
⌧
A-13
Dynamically access additional
spectrum.(see Section 4.3.2Error!
Reference source not found.).
Not clear what was causing problems,
but improved spectral efficiency
possible through cognitive techniques
could alleviate congestion.
Public Safety Special Interest Group
Use Cases for Cognitive Applications in Public Safety Communications Systems: Volume 1
SDRF-07-P-0019-V0.5
Time/
Event
PS Response
On-scene Problems /
Situation Awareness
Communication
Activity
Potential Cognitive Capability
therefore closed to the public –
outside the command and control
structure - at about noon, and
remained closed down until 4.45 pm.
During that period of time, O2
estimates that ‘Several hundred
thousand, possibly maybe even more
than a million’ attempted calls by
members of the public were lost. 19
Tavistock Square. The London
Ambulance Service manager at the
scene reported that the remainder of
the casualties still needed to go to
hospital.
12:12 PM
Russell Square. The scene was
clear of casualties.
Switch protocols to balance load or
allow additional access
As part of the analysis performed for
this document, it is noted that the other
protocols such as WiFi, WiMax were
unavailable so this use case is not
being addressed at this time.
1
19
Transcript of Committee meeting, 1 December 2005, Volume 2, page 88
A-14